New research tackles LLM agent challenges in edge-cloud, robustness, and reliability
ByPulseAugur Editorial·[294 sources]·
Recent research explores various aspects of Large Language Model (LLM) agents, focusing on improving their performance, robustness, and reliability. Studies introduce benchmarks like EC2-Bench for edge-cloud collaboration and ACED-Bench for causal reasoning. New methods such as TIPER and SkillFusion aim to enhance robustness and skill composition, while SchemaWeaver accelerates on-device agent execution. Additionally, research investigates agent unlearning to remove sensitive information and addresses issues like tool-result misbinding and semantic false consensus in multi-agent systems.
AI
IMPACT
These advancements aim to improve LLM agent capabilities in areas like edge computing, robustness, and causal reasoning, potentially leading to more reliable and efficient AI systems.
RANK_REASON
Multiple academic papers published on arXiv detailing new methods and benchmarks for LLM agents.
Most existing approaches to LLM swarm optimization are utility-driven, focusing on final task utility while leaving the processing, integration, and reliance of information within individual interactions largely implicit. To address this, we study LLM swarm interactions from an i…
Multi-agent systems driven by large language models (LLMs) rely on structured communication topologies to facilitate information exchange and collaborative reasoning. However, effectively scheduling these topologies to dynamically determine who talks to whom and when to communica…
arXiv:2610.22426v1 Announce Type: new Abstract: LLM agents are used to check records against an organization's requirements, for example whether a vulnerability report is ready for investigation. Requirements often allow several ways to be met, such as a severity rating with a cu…
arXiv:2610.22373v1 Announce Type: new Abstract: LLM-based agents are increasingly used as autonomous orchestrators in distributed and multi-agent systems, yet a control gap remains: no protocol-level mechanism allows human operators to inject binding directives that persist acros…
arXiv:2610.23342v1 Announce Type: new Abstract: Tool-augmented language-model agents rely on interface text: tool descriptions, argument schemas, and the candidate tool list shown at inference time. These interfaces often change after deployment even when the underlying tool sema…
arXiv cs.AI
TIER_1English(EN)·Oriol Miro-Lopez-Feliu, Oscar Mir\'o L\'opez-Feliu, Victor Gimenez-Abalos, Javier Vazquez-Salceda, Victor Carballo Araruna, Sergio Alvarez-Napagao·
arXiv:2610.22469v1 Announce Type: new Abstract: LLM-based agents are increasingly credited with the competences of practical reasoning: planning, monitoring, and correcting their own actions. Their success is attributed to the prompting strategies and agent architectures built ar…
arXiv:2609.32544v2 Announce Type: replace Abstract: In this paper, we use Pok\'emon Battles as a case study to investigate how to improve the performance of LLM-based agents in tasks that require opponent-aware planning without additional fine-tuning. We propose Long/Short-Term K…
arXiv cs.AI
TIER_1English(EN)·Trung-Kiet Huynh, Dao-Sy Duy-Minh, Chi-Nguyen Tran, Phu-Hoa Pham, Phu-Quy Nguyen-Lam, Le Hong Trang, The Anh Han·
arXiv:2610.22882v1 Announce Type: cross Abstract: Language-model agents are increasingly deployed to act on behalf of people in settings where individual decisions shape shared outcomes. We examine whether five production models adapt their contributions to changing risks and par…
arXiv:2610.22684v1 Announce Type: cross Abstract: Forecasting an unresolved event requires finding relevant evidence and deciding how strongly it supports each outcome. Yet outcome-based training of Large Language Model (LLM) forecasters often begins with preassembled evidence at…
arXiv:2610.23071v1 Announce Type: new Abstract: Large language models increasingly act as agents that call external tools, yet most tool-safety research examines only what follows a tool call. The prior decision of whether to call the tool remains largely unexamined and can influ…
arXiv:2609.34971v2 Announce Type: replace Abstract: Large Language Models (LLMs) have shown strong performance on tool-use agentic tasks when given a fixed tool schema. Yet a tool schema is not the action space of an agent; it is merely one interface representation of it. The sam…
As model reasoning gets cheaper, the physical experiment becomes the slow and costly step in discovery. Each experiment returns a noisy reading: a high value can mean a good candidate or a lucky draw. With a fixed budget, an AI agent must keep choosing between exploring a new can…
Large language model (LLM) multi-agent systems (MASs) solve complex tasks through interacting agents, yet their reliability is difficult to analyze because failures depend on both individual agents and how information and errors propagate across the system. Existing approaches la…
How does post-training change cooperation during repeated interaction? Within each of seven released model lineages, we compare pretrained (Base), supervised fine-tuned (SFT), and subsequent preference- or reward-optimized (Aligned) checkpoints across seven social dilemmas. Copie…
arXiv cs.AI
TIER_1English(EN)·Kshitij Mishra, Abhijith Sharma, Nils Lukas, Salem Lahlou·
arXiv:2610.22124v1 Announce Type: new Abstract: Modern agentic systems extend language models with external tools, enabling them to complete complex tasks through multi-turn interactions. As these agents are deployed in consequential settings, they are often constrained by polici…
arXiv:2610.21624v1 Announce Type: new Abstract: Real-world tasks are often complex and noisy, posing substantial challenges for LLM agents. Yet existing evaluations lack a systematic framework for jointly characterizing potential perturbation targets and types. We use this two-di…
arXiv:2610.21604v1 Announce Type: new Abstract: Scientific inquiry involves iterative cycles of exploration, hypothesis generation, and reasoning over observations to uncover relationships in complex systems. Large language models (LLMs) have demonstrated strong capabilities in s…
arXiv:2610.21405v1 Announce Type: new Abstract: Long-term memory has become a crucial component of large language model (LLM) agents, with most research focusing on retrieving relevant information accurately and efficiently. However, a dependable memory system must also forget se…
arXiv:2610.22273v1 Announce Type: cross Abstract: Reasoning over long document contexts remains a challenge for large language models (LLMs). Standard retrieval methods are inherently reactive, bottlenecking agents with complex, cross-document reasoning at inference time, while p…
arXiv cs.AI
TIER_1English(EN)·Cesare Zavattari, Alessandro Tommasi, Giuseppe Prencipe·
arXiv:2610.21708v1 Announce Type: cross Abstract: When an LLM agent misbehaves, a human steps in, and that intervention is usually discarded or replayed as training data. We treat it instead as a measurement: a label pointing at the internal computation that failed. We study it i…
arXiv:2610.21267v1 Announce Type: cross Abstract: Modern agents select tools from registries, but exposing entire registries incurs high token costs and can impair selection as they grow. We present TRACER, a retrieval front-end that indexes observed tool-call sequences in a trie…
arXiv cs.AI
TIER_1Nederlands(NL)·Liangjie Huang, Henry Peng Zou, Wei-Chieh Huang, Wooseong Yang, Yuwei Han, Philip S. Yu·
arXiv:2610.22138v1 Announce Type: new Abstract: LLM systems increasingly rely on groups of agents that discuss, critique, and check one another, assuming that peers will catch individual errors. We test this assumption against a minimal adversary. In a ten-agent factual discussio…
arXiv:2610.22031v1 Announce Type: new Abstract: Behavioral consistency across repeated executions provides an informative signal for assessing the reliability of large language model (LLM) agents. However, existing approaches typically summarize consistency with a single task-lev…
arXiv:2610.21218v1 Announce Type: new Abstract: Agentic large language models have been documented to lose reasoning capability when placed inside a harness, the software that mediates their observations and actions. To reconcile these losses, we introduce AutoOnboarder, an auton…
arXiv cs.LG
TIER_1English(EN)·Kseniia Strelbytska, Bal\'azs Szil\'agyi, Anna M\'esz\'aros, Ferenc Husz\'ar·
arXiv:2610.14923v2 Announce Type: replace Abstract: The TLA+ (temporal logic of actions) formal specification toolkit has seen widespread adoption in modeling modern distributed software systems including databases, storage systems, ledgers and coordination protocols. This popula…
arXiv:2602.18891v2 Announce Type: replace-cross Abstract: Advances in large language models (LLMs) are rapidly transforming scientific work, yet empirical evidence on how these systems reshape research activities remains limited. We report a mixed-methods pilot evaluation of an A…
arXiv:2610.04196v2 Announce Type: replace Abstract: Synchronous policy updates can limit the efficiency of post-training for long-horizon LLM agents. When trajectories span multiple interaction turns, completion times can vary substantially, and synchronous update barriers leave …
arXiv cs.AI
TIER_1English(EN)·Leyang Shen, Yaxin Gong, Yang Zhang, Chenxiao Fan, Jiakai Wang, Dong Wang, Yang Liu, Fuli Feng, Chun Kai Ling, Tat-Seng Chua·
arXiv:2610.22316v1 Announce Type: cross Abstract: Large language model (LLM)-based agents and multi-agent systems increasingly tackle complex tasks through long-horizon tool use and inter-agent collaboration. However, these systems are prone to deviating from the overall goal as …
arXiv cs.MA (Multiagent)
TIER_1English(EN)·The Anh Han·
Language-model agents are increasingly deployed to act on behalf of people in settings where individual decisions shape shared outcomes. We examine whether five production models adapt their contributions to changing risks and partner behaviour, or instead follow a relatively fix…
Large language model (LLM)-based agents and multi-agent systems increasingly tackle complex tasks through long-horizon tool use and inter-agent collaboration. However, these systems are prone to deviating from the overall goal as execution unfolds, which wastes execution resource…
arXiv cs.MA (Multiagent)
TIER_1Nederlands(NL)·Philip S. Yu·
LLM systems increasingly rely on groups of agents that discuss, critique, and check one another, assuming that peers will catch individual errors. We test this assumption against a minimal adversary. In a ten-agent factual discussion, one or three adversarial agents repeatedly as…
Behavioral consistency across repeated executions provides an informative signal for assessing the reliability of large language model (LLM) agents. However, existing approaches typically summarize consistency with a single task-level score, without identifying where behavioral d…
Modern agents select tools from registries, but exposing entire registries incurs high token costs and can impair selection as they grow. We present TRACER, a retrieval front-end that indexes observed tool-call sequences in a trie and ranks continuations using a learned blend of …
arXiv:2610.17436v1 Announce Type: new Abstract: Large language models (LLMs) on mobile devices can act as agents that call external software tools to complete complex tasks, but the major challenge of supporting on-device mobile agentic AI with LLMs is the long latency of LLM too…
arXiv cs.AI
TIER_1English(EN)·Yiwei Chen, Amila Weerasinghe, Siddharth Biswal, Utkrisht Rajkumar, Olcay Boz, Taylor Berg-Kirkpatrick, Rose Yu·
arXiv:2610.14381v1 Announce Type: new Abstract: Large language models are increasingly used as scientific assistants, yet we still lack reliable ways to evaluate whether they can investigate a system rather than merely answer a prompt. A useful evaluation must let agents choose e…
arXiv:2610.15481v1 Announce Type: new Abstract: Independently authored skills can impose incompatible requirements, competing preferences, and redundant instructions when jointly loaded by an LLM agent. We propose SkillFusion, an offline compiler that reconciles these interaction…
arXiv:2610.16155v1 Announce Type: new Abstract: Interactive-agent evaluations commonly rely on standardized LLM-based user simulators that exhibit broadly cooperative behavior, potentially overlooking failures that arise under diverse but realistic user interactions. This can lea…
arXiv cs.AI
TIER_1English(EN)·Keru Chen, Sen Lin, Yingbin Liang, Nathaniel D. Bastian, Shaofeng Zou·
arXiv:2610.17893v1 Announce Type: new Abstract: Effective collaboration among large language model (LLM) agents in multi-agent systems requires sharing relevant evidence, requesting missing information, and incorporating teammates' contributions into decisions. Training these beh…
arXiv:2610.17869v1 Announce Type: new Abstract: Multi-agent Large Language Model (LLM) systems typically commit to a deliberation protocol before inference. A fixed protocol cannot allocate computation differently to resolved and unresolved instances, and it may mistake repeated …
arXiv:2610.17816v1 Announce Type: new Abstract: Edge-cloud collaboration combines the lower inference overhead and greater data locality of edge-side execution with the stronger capabilities of cloud-hosted models. However, as LLM agents increasingly operate in long-horizon scena…
arXiv cs.AI
TIER_1English(EN)·Chun-Yu Hsu, Lun-Wei Ku, Chung-Chi Chen·
arXiv:2610.17723v1 Announce Type: new Abstract: People act on evaluative judgments produced by LLM agents, and in multi-agent systems those judgments pass through several agents before reaching a person. We show that a workflow's information architecture---source access, handoff …
arXiv:2610.17659v1 Announce Type: new Abstract: Large language model (LLM) agents have shown strong capabilities in interactive problem solving, tool use, and long-horizon task execution. However, they may also acquire sensitive or private knowledge during training and expose it …
arXiv cs.AI
TIER_1English(EN)·Marcos Lima Romero, Ricardo Suyama·
arXiv:2610.17118v1 Announce Type: new Abstract: Future AI-native networks are being designed to apply autonomous agents built on large language models (LLMs) in network management through intents. When intents from different domains meet on the same resource, an agent must recogn…
arXiv:2610.16333v1 Announce Type: new Abstract: As AI agents increasingly act through external tools, tool results are no longer simple outputs but execution artifacts that must remain correctly associated with the context that produced them. Yet as results persist across caches,…
arXiv:2610.19712v1 Announce Type: new Abstract: Multi-session agents can switch between LLMs to balance performance and inference cost. Such switches require transferring a history for continuity, yet their effects on the recipient remain underexplored. Through controlled experim…
arXiv cs.CL
TIER_1English(EN)·Md Tahmid Rahman Laskar, Gundeep Singh, Xue-Yong Fu, Karol Chang, Kevin Sanders, Shi Zong, Tania Habib, Julien Bouvier Tremblay, Shayna Gardiner, Harsh Saini, Matthias Lee, Elena Khasanova, Quinten McNamara, Shashi Bhushan TN·
arXiv:2610.14346v2 Announce Type: replace Abstract: Enterprise customer-support agents increasingly rely on Large Language Models (LLMs) to execute bounded, tool-using workflows. We study this problem through the lens of Skill LLMs: specialized executors that operate within a sel…
arXiv:2610.20576v1 Announce Type: cross Abstract: Post-training binds the competence of an LLM agent to the interface it was trained on. Semantics-preserving perturbations of that interface, such as renaming an action or changing the notation, leave the underlying task unchanged …
arXiv cs.CL
TIER_1English(EN)·Alexander Gill, Guram Gogia, Francisco Guzm\'an·
arXiv:2610.19614v1 Announce Type: cross Abstract: In this work, we reframe agentic trajectories as paths through a shared embedding space of actions and environmental observations. We show that successful and failed rollouts of the same task separate in this space, with simple pe…
arXiv:2610.19560v1 Announce Type: cross Abstract: In long-horizon tool-calling tasks, LLM agents can fail not only because of tool-use errors, but also because of ineffective interaction with users. Yet existing benchmarks largely assume cooperative and behaviorally stable users,…
arXiv cs.CL
TIER_1English(EN)·Delip Rao, Ahmed Muharram, Chris Callison-Burch·
arXiv:2610.20457v1 Announce Type: new Abstract: Co-scientists and agentic scientific claim verifiers are increasingly used to assess claims across the sciences. Natural-language scientific claims, written for human experts, are often underspecified and so admit several defensible…
arXiv:2610.19659v1 Announce Type: new Abstract: People delegate multi-step work to LLM agents and read only the reply that the agent hands back. Current evaluation scores task completion and the faithfulness of reports, and it leaves open whether the requester can understand and …
arXiv:2610.19367v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for multi-turn agentic tasks that require interactions with users and environments over extended sequences of actions. A common approach to improving LLM agents is to use more capab…
arXiv cs.AI
TIER_1English(EN)·Ali Afoud, Jie JW Wu·
arXiv:2609.31937v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents built on the ReAct pattern concentrate four responsibilities in one model: selecting a strategy, choosing each action, formatting it, and judging whether the result is adequate. Nothing ou…
arXiv:2508.04903v4 Announce Type: replace-cross Abstract: Large language model (LLM) agents are increasingly organized into multi-agent systems for complex reasoning and decision-making tasks. However, scaling such systems can substantially increase inference cost, especially whe…
arXiv:2608.28399v2 Announce Type: replace Abstract: In financial markets, a sequential policy that reacts systematically to price movements may become predictable to other market participants. This paper studies whether large language model (LLM) agents exhibit such directional s…
arXiv:2607.28103v2 Announce Type: replace Abstract: Memory-augmented LLM-based agents are vulnerable to memory injection attacks: Agents may retrieve poisoned memory from attackers, which diverts their behavior from initial user intent and finally causes task failure. However, ex…
arXiv:2606.22953v2 Announce Type: replace Abstract: Context managers summarize or evict old messages so agents can operate beyond a finite window. This creates a reliability problem: removing history that still shapes the model's computation may change later actions, while retain…
arXiv:2606.22936v2 Announce Type: replace Abstract: Tool-using LLM agents make early choices about what to search for, which evidence to trust, and what to do next. A wrong early choice can steer every later tool call, so an agent may produce a coherent but incorrect result. Exis…
arXiv cs.AI
TIER_1English(EN)·James Mooney, Josef Woldense, Zheng Robert Jia, Shirley Anugrah Hayati, My Ha Nguyen, Vipul Raheja, Dongyeop Kang·
arXiv:2509.03736v3 Announce Type: replace Abstract: LLM agents are increasingly proposed as substitutes for human participants in social research, usually on the strength of survey responses that match those of the humans they represent. We argue that such output-level alignment …
arXiv:2610.17544v1 Announce Type: cross Abstract: Thought templates, skill libraries, agent memories and caches turn many reasoning steps into one. We call this reasoning contraction. A contracted memory saves reasoning steps and errors until its facts change, then quietly answer…
arXiv:2610.16489v1 Announce Type: cross Abstract: Reinforcement learning (RL) can substantially improve the reasoning capabilities of large language models, but an effective training run requires considerable resources. Reusing the computation acquired through RL across models wo…
arXiv:2610.16067v1 Announce Type: cross Abstract: Agent evaluation harnesses are software systems that run test suites, generate inputs, and decide outcomes with test oracles. This paper audits such software for reliability defects that score tables cannot reveal. We introduce an…
arXiv cs.AI
TIER_1English(EN)·Sunday Oyinlola Ogundoyin, Muhammad Ikram·
arXiv:2610.16039v1 Announce Type: cross Abstract: Multi-agent LLM systems (MALS) introduce a privacy-leakage problem that conventional sensitivity-based defenses miss, since information disclosed to one authorized agent can propagate through routine coordination to an unauthorize…
arXiv cs.AI
TIER_1English(EN)·Kaijie Zhu, Weixiang Yan, Omar Attia, Abbas Kazerouni, Alexander Toshev, Andrew Szot·
arXiv:2610.15840v1 Announce Type: cross Abstract: Reinforcement learning (RL) for large language model (LLM) agents remains challenging in long-horizon tasks with sparse outcome rewards, where assigning credit to individual actions is difficult. Methods such as GRPO assign the sa…
arXiv:2610.15513v1 Announce Type: cross Abstract: Safety-aligned large language models (LLMs) remain vulnerable to adversarial prompts that bypass their safety training. Existing automated red-teaming methods either require privileged access unavailable in production, rely on man…
arXiv:2610.15459v1 Announce Type: cross Abstract: LLM-based multi-agent systems (LLM-MAS) use specialized agents that exchange messages to accomplish complex tasks. Their reliance on networked protocols, remote tools, and model endpoints exposes communication to message loss, del…
arXiv:2610.14238v1 Announce Type: cross Abstract: LLM data engineering requires iterative experimentation to determine how data selection, rewards, and curricula affect model learning, making it a natural setting for AI-for-AI (AI4AI) research. Can agents use experimental feedbac…
arXiv:2610.14000v1 Announce Type: cross Abstract: Group-relative reinforcement learning methods such as GRPO train large language model (LLM) agents by comparing rollouts for the same task. When every trajectory in a group fails with the same reward, relative advantages collapse …
arXiv:2610.18780v1 Announce Type: new Abstract: Large language model multi-agent systems extend individual agents' capabilities through collaboration across specialized roles, tools, and workflows. Yet collaboration makes failures harder to diagnose: one agent may misinterpret a …
arXiv:2610.18614v1 Announce Type: new Abstract: Controlling an acting LLM agent means deciding, at each step of a long-horizon task, whether to trust its next action, abstain, or hand off to a human. That decision needs a per-step confidence signal predicting whether the current …
arXiv:2610.18572v1 Announce Type: new Abstract: Automated post-training transforms pre-trained large language models (LLMs) into models capable of handling diverse domains by automatically constructing and optimizing a recipe of coupled decisions over data sources, data processin…
arXiv:2610.18522v1 Announce Type: new Abstract: Reusable skills let LLM agents load task-specific instructions on demand, but they make execution depend on implicit routing over skill metadata. We test whether this design scales by holding SkillsBench tasks and their golden skill…
arXiv:2610.18378v1 Announce Type: new Abstract: Applying process-based Earth science models to new sites and questions often requires expert calibration decisions. Reliable delegation to language-model agents must preserve scientific competence and reproducibility. Here we develo…
arXiv cs.AI
TIER_1English(EN)·Arastun Mammadli, Herbert Woisetschl\"ager, Jawad Fayaz, Shiqiang Wang·
arXiv:2610.18354v1 Announce Type: new Abstract: Game-theoretic benchmarks claim that large language model (LLM) strategic decision-making varies across game types, interaction structures, and capability dimensions. However, they only evaluate strategic performance across individu…
Librarian Agents use reinforcement learning to train builder agents that reorganize document corpora into reusable hierarchical workspaces, improving downstream reasoning accuracy and amortizing costs across queries.
Repeated event forecasting requires extracting predictive signals from noisy evidence and evaluating their value across successive decisions. We introduce JockeyBench, a benchmark and evaluation protocol that studies this problem through historical Japanese horse racing. The prot…
Large language model (LLM) agents can reuse experience from prior task executions to improve performance on subsequent tasks. Although existing methods enable cross-agent experience sharing, differences in agents' execution histories and task requirements make shared experiences …
In long-horizon tool-calling tasks, LLM agents can fail not only because of tool-use errors, but also because of ineffective interaction with users. Yet existing benchmarks largely assume cooperative and behaviorally stable users, leaving the effect of emotional variation on agen…
Large language models (LLMs) are increasingly used for multi-turn agentic tasks that require interactions with users and environments over extended sequences of actions. A common approach to improving LLM agents is to use more capable expert models to provide training supervision…
AgentSafeEvo uses adversarial co-evolution to train LLM agents to resist malicious tool-use attacks across diverse tasks without sacrificing general capabilities.
Reusable skills let LLM agents load task-specific instructions on demand, but they make execution depend on implicit routing over skill metadata. We test whether this design scales by holding SkillsBench tasks and their golden skills fixed while adding irrelevant skills. We find …
Game-theoretic benchmarks claim that large language model (LLM) strategic decision-making varies across game types, interaction structures, and capability dimensions. However, they only evaluate strategic performance across individual fixed models. In parallel, contemporary LLM r…
Effective collaboration among large language model (LLM) agents in multi-agent systems requires sharing relevant evidence, requesting missing information, and incorporating teammates' contributions into decisions. Training these behaviors is challenging because final task outcome…
Effective collaboration among large language model (LLM) agents in multi-agent systems requires sharing relevant evidence, requesting missing information, and incorporating teammates' contributions into decisions. Training these behaviors is challenging because final task outcome…
Multi-agent Large Language Model (LLM) systems typically commit to a deliberation protocol before inference. A fixed protocol cannot allocate computation differently to resolved and unresolved instances, and it may mistake repeated support for independent confirmation. We study t…
Multi-agent Large Language Model (LLM) systems typically commit to a deliberation protocol before inference. A fixed protocol cannot allocate computation differently to resolved and unresolved instances, and it may mistake repeated support for independent confirmation. We study t…
Controlling an acting LLM agent means deciding, at each step of a long-horizon task, whether to trust its next action, abstain, or hand off to a human. That decision needs a per-step confidence signal predicting whether the current rollout will eventually succeed. The agent's own…
Strategic model selection for game-theoretic LLM evaluation reveals that multi-armed bandit methods outperform embedding-based selectors but remain below fixed-model baselines against adaptive opponents.
In LLM agent organizations, shared records determine which facts are available for later decisions. A sender consults interaction histories before deciding whether to report an event. A delivery gap across histories does not show whose relationship drove the choice or whether the…
arXiv cs.AI
TIER_1English(EN)·Yongjian Tang, Costel Moraru, Doruk Tuncel, Hanya Elhashemy, Steffen Klepke, Thomas Runkler·
arXiv:2610.13273v1 Announce Type: cross Abstract: API specifications serve as critical documentation for modern software development, directly impacting integration efficiency and client code generation. While recent advancements have leveraged Large Language Models (LLMs) to gen…
arXiv:2610.12789v1 Announce Type: new Abstract: Self-evolving LLM agents typically improve how they act through accumulated experience, reflection, or skills. Yet even procedurally capable agents can fail when their beliefs about environment dynamics are incomplete, incorrect, or…
arXiv:2610.12828v1 Announce Type: new Abstract: Outcome-supervised group-relative policy optimization (GRPO) assigns the same advantage to every token in an LLM agent trajectory, leaving useful and unhelpful decisions undifferentiated. Hindsight credit assignment (HCA) can refine…
arXiv cs.LG
TIER_1English(EN)·Ning Liu, Ian Li, Kristina Klinkner·
arXiv:2610.12651v1 Announce Type: new Abstract: LLM agents executing multi-step workflows fail in two fundamentally different ways: recoverable slips, which further attempts can repair, and knowledge gaps, which retrying cannot resolve. Expert feedback can address the latter, but…
arXiv:2605.07251v2 Announce Type: replace Abstract: Large Language Models (LLMs) have become increasingly capable as tool-using agents, with benchmarks spanning diverse general agentic tasks. Yet rigorous evaluation of scientific tool use remains limited. In chemistry, recent age…
arXiv:2610.13755v1 Announce Type: cross Abstract: Training long-horizon language model agents requires assigning credit to intermediate decisions from sparse terminal feedback. Within a finite group of sibling rollouts, useful action comparisons depend on their context, evidence …
arXiv cs.AI
TIER_1English(EN)·Ivan Moloshnikov, Dmitry Kunitsyn, Alexander Sboev, Maksim Skorokhodov, Roman Rybka, Yuanchao Liu·
arXiv:2610.13663v1 Announce Type: new Abstract: Many machine learning approaches are two-level: an expert selects a class of methods, within this class the choice of key components is critical and usually manual. This paper addresses the automatic tuning of these components and p…
arXiv:2610.12975v1 Announce Type: new Abstract: We present SmartSalesBench, the first multi-turn conversational benchmark explicitly designed to evaluate large language model (LLM) sales agents against psychologically grounded, resistant buyer personas. Existing task-oriented dia…
arXiv:2610.13144v1 Announce Type: new Abstract: LLM-based multi-agent systems (MAS) have emerged as a promising paradigm for enhancing complex reasoning through role specialization and collaboration, yet a key challenge remains: how can collaboration be adapted to a backbone's co…
arXiv:2610.13335v1 Announce Type: new Abstract: Tool-using agents increasingly require a common substrate for building, training, evaluating, distilling, and maintaining agentic systems, yet existing tooling typically addresses these stages in isolation. We present AgentTune, a s…
arXiv:2610.13064v1 Announce Type: new Abstract: Large Language Model (LLM) agents have demonstrated remarkable capabilities in tackling complex tasks through agentic workflows. However, manually designing these workflows is labor-intensive and lacks flexibility to adapt to divers…
Safety-aligned large language models (LLMs) remain vulnerable to adversarial prompts that bypass their safety training. Existing automated red-teaming methods either require privileged access unavailable in production, rely on manually engineered per-model templates, or optimize …
arXiv cs.MA (Multiagent)
TIER_1English(EN)·George J. Pappas·
Natural language specifications let non-expert users task robot teams, and large language models (LLMs) have recently proven effective at grounding such semantic specifications into plans for heterogeneous teams. Existing approaches assume missions that terminate and environments…
LLM agents increasingly divide enterprise work across roles with different goals, authority, access, and information. Because each agent often sees only a local snapshot of the enterprise workflow, locally reasonable actions can still violate organization-wide constraints (e.g., …
Many machine learning approaches are two-level: an expert selects a class of methods, within this class the choice of key components is critical and usually manual. This paper addresses the automatic tuning of these components and proposes a framework transferable across models o…
API specifications serve as critical documentation for modern software development, directly impacting integration efficiency and client code generation. While recent advancements have leveraged Large Language Models (LLMs) to generate initial specifications from source code, exi…
arXiv:2610.12375v1 Announce Type: new Abstract: Agents are deployed in applications from trip planners and stock trading to IT incident triage. In most cases, LLM agents work autonomously with minimal rule-based safeguarding, leading to cost and safety issues from irreversible ac…
arXiv cs.CL
TIER_1English(EN)·Isadora White, Yasaman Jafari, Taylor Berg-Kirkpatrick·
arXiv:2606.22698v2 Announce Type: replace-cross Abstract: As LLM-powered scams proliferate, black-box forensics for conversational LLM agents offers a path to accountability for systems hidden behind anonymous endpoints. Identifying the base model behind a chatbot endpoint (attri…
arXiv:2603.08993v2 Announce Type: replace-cross Abstract: System prompts for LLM-based coding agents are software artifacts that govern agent behavior, yet lack the testing infrastructure applied to conventional software. We present Arbiter, a framework combining formal evaluatio…
arXiv:2610.12183v1 Announce Type: cross Abstract: Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensive and limited. Recent large language model (LLM) agents offer a new way to approach BBO by combining task sema…
arXiv cs.AI
TIER_1English(EN)·Irene Weber (University of Applied Sciences Kempten, Germany)·
arXiv:2610.11899v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly embedded as components in software systems, marketed under labels such as chatbot, copilot, retrieval-augmented generation, workflow, coding agent and AI agent. Whether these labels de…
arXiv cs.AI
TIER_1English(EN)·Masaaki Nakatsu (AO, Inc. / OrbLabs AG), Reno Wang (AO, Inc.)·
arXiv:2610.11542v1 Announce Type: cross Abstract: Multi-agent LLM systems negotiating with a stateful counterpart waste model calls in three ways: polite loops that never meet the counterpart's hidden acceptance condition, malformed outputs that trigger retries, and compliance de…
arXiv cs.AI
TIER_1English(EN)·Jorge Garc\'ia-Carrasco, Sergio Garc\'ia-Carrasco, Alejandro Mat\'e, Juan Trujillo·
arXiv:2610.11447v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed as coding agents that edit files, run builds and tests, inspect execution results, and repair software iteratively. Embedded firmware is a demanding target because correctness…
arXiv cs.AI
TIER_1English(EN)·Min-Young Yu, Tony Kim, Jang Won Choi·
arXiv:2610.11030v1 Announce Type: cross Abstract: Tool-using LLM agents violate the policies they are deployed to enforce, often silently. Prior defenses hand-write rules, query an LLM verifier per action, or compile policies through heavyweight formal machinery. Naive compilatio…
arXiv:2610.11008v1 Announce Type: cross Abstract: Language-model agents favor their own group because they have watched their members favor each other. The group label alone does little once the decision has a cost; what drives favoritism is observed behavior, and an individual's…
arXiv:2610.11600v1 Announce Type: new Abstract: LLM-based multi-agent systems (MASs) are increasingly used to solve complex tasks through coordinated reasoning, tool use, and interaction with external resources. However, attributing failures in such systems remains challenging be…
arXiv:2610.11334v1 Announce Type: new Abstract: In LLM-based agent systems, failures can originate from early steps whose effects propagate through subsequent interactions, making their origins difficult to identify. To trace such failures back to their origin, failure attributio…
arXiv:2610.11007v1 Announce Type: new Abstract: At the start of every session, LLM agents load a fixed context file, such as $\texttt{AGENTS.md}$. Each loaded token in the file is charged again in every later round of the session, and these files can degrade performance as they g…
arXiv:2610.10629v1 Announce Type: new Abstract: Self-improving LLM agents can adapt a credit pipeline to a changed rule, but an agent that rewrites itself destroys the artefact a supervisor reviews: a named change, a recorded test, an approval. We argue that self-evolution is rev…
Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensive and limited. Recent large language model (LLM) agents offer a new way to approach BBO by combining task semantics, computation, optimization tools, and feedba…
LLM-based multi-agent systems (MASs) are increasingly used to solve complex tasks through coordinated reasoning, tool use, and interaction with external resources. However, attributing failures in such systems remains challenging because the observed outcome often does not direct…
arXiv:2610.09872v1 Announce Type: cross Abstract: Evaluating agents by outcomes alone can obscure the capabilities that produce them. This problem is especially pronounced in evolving environments, where outcomes reflect a closed-loop interaction between agent behavior and changi…
arXiv:2605.12894v2 Announce Type: replace-cross Abstract: Large Language Model (LLM) agents are increasingly deployed in settings where they interact with diverse users, including those who are unclear, impatient, or reluctant to share information. However, collecting real intera…
arXiv:2610.06401v2 Announce Type: replace-cross Abstract: Tool-using language-model agents are vulnerable to indirect prompt injection because they must act on untrusted external content. Existing training-time defenses can reduce attack success rates, but often at the cost of ge…
arXiv cs.CL
TIER_1English(EN)·Sergei Polezhaev, Barys Liskavets, Ori Press, Alexander Golubev·
arXiv:2610.09858v1 Announce Type: new Abstract: Large language model agents tackle multi-step tasks by interleaving reasoning and tool calls with observations from the environment. Prior work has shown that natural-language feedback can help these agents revise their decisions du…
arXiv:2610.09115v1 Announce Type: cross Abstract: Steering an LLM agent means deciding whether to correct it, at which step, and with which mechanism. Uncertainty is often used to decide when to correct an agent, but whether it can guide these decisions remains unclear. We steer …
arXiv:2610.09624v1 Announce Type: cross Abstract: Tool calling, invoking external tools on demand, is central to agentic LLMs, yet the mechanism that decides whether a model calls a tool or responds directly remains poorly understood. Agentic prompts are long and heavily scaffold…
arXiv:2603.06859v4 Announce Type: replace Abstract: Credit assignment for a team of LLM agents, what each message was worth, has mostly been treated as prediction: a learned critic, a trajectory-level score, or an agent ablation stands in for the counterfactual that defines credi…
arXiv cs.CL
TIER_1English(EN)·Masaaki Nakatsu (AO, Inc. / OrbLabs AG), Reno Wang (AO, Inc.)·
arXiv:2610.09772v1 Announce Type: new Abstract: Small language-model agents on edge devices must hold a persona and reason correctly at once, inside one context window that fills with conversational history and persona instructions. We study what happens to the logical part of su…
Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for understanding how effectively agents acquire and use new knowledge. Existing benchmarks have sought to evaluate this ability, b…
Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensive and limited. Recent large language model (LLM) agents offer a new way to approach BBO by combining task semantics, computation, optimization tools, and feedba…
Language-model agents favor their own group because they have watched their members favor each other. The group label alone does little once the decision has a cost; what drives favoritism is observed behavior, and an individual's own record can override it. We test this in small…
Self-improving LLM agents can adapt a credit pipeline to a changed rule, but an agent that rewrites itself destroys the artefact a supervisor reviews: a named change, a recorded test, an approval. We argue that self-evolution is reviewable only if it is confined to the runtime ha…
Evaluating agents by outcomes alone can obscure the capabilities that produce them. This problem is especially pronounced in evolving environments, where outcomes reflect a closed-loop interaction between agent behavior and changing external conditions. We introduce LiveMACEBench…
arXiv cs.AI
TIER_1English(EN)·Wei Shi, Ziheng Peng, Sihang Li, Xiting Wang, Xiang Wang, Mengnan Du, Na Zou·
arXiv:2605.18882v2 Announce Type: replace-cross Abstract: LLM agents exhibit a consistent tendency to over-call, invoking tools even in situations where none is needed. On the When2Call benchmark, six models from three families show high call accuracy but much lower no-call accur…
arXiv:2610.07967v1 Announce Type: new Abstract: As large language model (LLM) agents become increasingly autonomous, they may pursue task performance through deception, raising concerns about their reliable deployment. Existing evaluations show that LLM agents can deceive, but of…
arXiv:2610.08402v1 Announce Type: new Abstract: Multi-turn LLM agents often receive sparse task feedback across several interactions, while generating each response token by token. This creates two related credit-assignment questions: which responses helped achieve the outcome, a…
arXiv:2610.07359v1 Announce Type: new Abstract: Runtime gates for agent tool calls are stacked on the assumption that their errors multiply. We test it on 1,119 labelled agent actions from three corpora, without an adaptive adversary. The stack has one deterministic rule layer an…
arXiv cs.AI
TIER_1English(EN)·Toby D. Pilditch, Konstantinos Voudouris, Alexandra Abbas, Cozmin Ududec·
arXiv:2610.08364v1 Announce Type: new Abstract: Frontier AI evaluations increasingly use open-ended, agentic, long-horizon tasks whose transcripts can span hundreds of pages of outputs and actions from complex multi-agent networks. The observability envelop-the range of what eval…
arXiv cs.AI
TIER_1English(EN)·Ankit Sonthalia, Haritz Puerto, Alexander Rubinstein, Martin Gubri, Seong Joon Oh·
arXiv:2610.08775v1 Announce Type: new Abstract: Large language models (LLMs) can solve many narrow tasks, but querying them separately for millions of related instances can be prohibitively expensive. Can LLM agents autonomously create cheaper solutions for such workloads? We cal…
arXiv cs.AI
TIER_1English(EN)·Xinran Zheng, Xin Fan Guo, Zhiqiang Hao, Fan Yang, Xingzhi Qian, Jiawei Du, Jinfeng Xu, Zheng Xing, Shuo Yang, Xingjun Wang·
arXiv:2610.06966v1 Announce Type: cross Abstract: Indirect prompt injection (IPI) hides adversarial instructions in content that large language model (LLM) agents read at runtime. As agents compose heterogeneous capability units, including Tools, MCP servers, and Skills, the carr…
arXiv cs.AI
TIER_1English(EN)·Heewon Park, Somin Im, Minhae Kwon·
arXiv:2610.07335v1 Announce Type: cross Abstract: Improving the reliability of large language model (LLM) agents in long-horizon decision-making remains a key challenge. When deployed as autonomous agents interacting with complex environments, early mistakes can propagate through…
arXiv:2606.21399v2 Announce Type: replace Abstract: Runtime oversight often intervenes when an LLM agent's calibrated failure score crosses a threshold. Yet states with the same failure risk can differ in whether intervention helps. Strictly increasing recalibration preserves the…
arXiv:2610.07948v1 Announce Type: new Abstract: When using an LLM agent in a consequential domain, making an informed decision about whether to trust its output or intervene requires calibrated confidence in the agent's success. Confidence estimation for agents is difficult becau…
arXiv:2610.08082v1 Announce Type: new Abstract: LLM tool-use agents operate in dynamic environments where many actions carry operational risk. However, most safety mechanisms react only after errors manifest. Existing pre-emptive approaches either fine-tune the agent on chain-of-…
arXiv:2610.08215v1 Announce Type: new Abstract: Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for understanding how effectively agents acquire and use new knowledge. Existing be…
LiveMACEBench evaluates persistent LLM agents in live financial markets using decision traces to reveal gaps between outcomes and underlying capabilities.
LLM tool-use agents operate in dynamic environments where many actions carry operational risk. However, most safety mechanisms react only after errors manifest. Existing pre-emptive approaches either fine-tune the agent on chain-of-thought deliberation or compile natural-language…
As large language model (LLM) agents become increasingly autonomous, they may pursue task performance through deception, raising concerns about their reliable deployment. Existing evaluations show that LLM agents can deceive, but often examine isolated scenarios or narrowly defin…
In decentralized consumer-to-consumer (C2C) marketplaces, people list goods, negotiate with strangers, and rate one another, so trust rests on reputation. Large language model (LLM) agents now act for users, raising risks to their money, privacy, and reputation. We introduce Baza…
Recent work on multi-agent LLM systems reaches sharply different conclusions: some results show that a single agent with the same information and compute should dominate a delegated system, others that multi-agent gains grow with task depth. We argue that much of the disagreement…
arXiv cs.AI
TIER_1English(EN)·Changxiu Ji, Amy Lu, Qizheng Zhang, Kunle Olukotun·
arXiv:2610.02994v1 Announce Type: cross Abstract: LLM agents often fail mid-task due to invalid tool calls, repeated actions, or poorly grounded reasoning, and learning from these failures is a path to reliability. We find that how failure knowledge reaches the agent matters as m…
arXiv:2610.02702v1 Announce Type: new Abstract: Multi-agent debate is increasingly used to reach consensus among LLM agents, yet agents often yield to a unanimous majority. When an agent changes its answer, has it changed its mind or only its statement? We study this with two-hop…
arXiv:2610.03014v1 Announce Type: cross Abstract: Large language model (LLM)-based agents increasingly connect model-generated decisions to security-sensitive software capabilities such as command execution, filesystem access, network communication, browser control, and external …
arXiv cs.LG
TIER_1English(EN)·Zelin Zhao (Georgia Institute of Technology), Xinyu Guo (Georgia Institute of Technology), Jingyuan Zhang (Georgia Institute of Technology), Yuxuan Zhang (Etude AI), Yongxin Chen (Georgia Institute of Technology)·
arXiv:2610.02488v1 Announce Type: new Abstract: Language-model agents increasingly rely on harnesses that manage bounded context, persistent memory, tools, verification, and repeated execution, yet existing notions of model capability do not quantify the computational resources t…
arXiv:2610.03448v1 Announce Type: cross Abstract: LLM agents increasingly screen tool outputs with small prompt-injection detectors, and teams choose among detectors by their scores on public benchmarks. We ask whether those scores predict how a detector behaves inside an agent. …
arXiv:2610.03195v1 Announce Type: new Abstract: As LLM agents decide on users' behalf which product to buy, which hotel to book, or which paper to cite, a preference for items from certain sources (the sites or services they come from) shapes what users receive and which sources …
arXiv:2610.02425v1 Announce Type: new Abstract: Static evaluations credit a language model for naming the right move, but an agent must carry a plan through to a verified outcome while an opponent responds. We introduce XiangqiBench, an executable benchmark that measures this dif…
arXiv:2610.02267v1 Announce Type: new Abstract: Agent harnesses make many small, typed decisions per task: which model to call, which tool to use, whether retrieved text is relevant, whether an input carries an injection. System-1 decision models answer such questions in a single…
Developers can build LLM agents by adapting third-party models through benign post-training. We study a supply-chain threat in which an attacker supplies a model with a backdoor: hidden behavior that produces malicious outputs when a particular input pattern appears. Focusing on …
Layered trust models for multi-agent LLM systems remain largely conceptual: they name which dimensions of trust matter but not how layers combine, how their importance is set at runtime, or how trust should govern agent actions. This gap matters because higher inter-agent trust r…
AI agents based on foundation models (FMs) have demonstrated strong capabilities to perform complex open-ended tasks. However, they face some common challenges in practice: (a) agent behavior can deviate drastically even for semantically similar tasks, leading to catastrophically…
In August and September 2026, independent researchers publicly documented an unusual incident: thousands of autonomous agents, self identifying as OpenAI models on web research tasks, discovered and began using a small German wiki as an improvised message board posting roughly 18…
arXiv:2610.00511v1 Announce Type: new Abstract: Before a difficult decision, people often act simply to understand the situation better. We turn an object to see another side, place alternatives next to each other, or change one condition and observe what happens. These actions m…
arXiv:2610.00313v1 Announce Type: new Abstract: Scientific coding agents receive equations, boundary conditions, and output requirements in writing, then must assess the programs they revise. Rules to Tools (R2T) supplies prepared executable checks of public scientific requiremen…
arXiv:2610.00012v1 Announce Type: new Abstract: LLM agents increasingly act through modular systems, such as order, payment, inventory, and shipment services, where actions in one module change which transitions are valid in another. Standard world models usually fit observationa…
arXiv cs.AI
TIER_1English(EN)·Shuyao Xiao, Shengling Wang, Xuan Chen, Ke Chao, Ming Cui, Feifei Qian, Chaoyang Mei, Fanlin Meng, Ziming Yu, Junxi Yin·
arXiv:2610.00372v1 Announce Type: new Abstract: Large language model agents rely on external harnesses to pass information between the model and its environment and to recover from execution errors. Yet recovery is usually judged only by average task success. This hides an import…
arXiv cs.AI
TIER_1English(EN)·Kay K\"ohle, Darko Anicic, Thomas A. Runkler, Ren\'e Graf·
arXiv:2610.01364v1 Announce Type: cross Abstract: Factories are shifting toward smaller lot sizes with high product customization, requiring frequent re-programming of flexible and reconfigurable automation systems. LLM-based agents can be deployed in two complementary roles: Off…
arXiv cs.AI
TIER_1English(EN)·Asad Ur Rehman, Syed Mohammad Kashif, Ruiyin Li, Peng Liang, Zengyang Li, Arif Ali Khan·
arXiv:2610.00905v1 Announce Type: cross Abstract: With the advancement of LLM-based multi-agent systems (MAS), an increasing number of opensource projects are adopting multi-agent architectures as the foundation of their core functionality. Although research and practice on MAS h…
arXiv:2610.01564v1 Announce Type: cross Abstract: LLM agents use skills to improve performance on specialized tasks. To complete a user request, an agent may invoke several skills in sequence, allowing information produced under one skill to guide the next. Because skills may com…
arXiv cs.AI
TIER_1English(EN)·Fengpeng Li, Qizhou Wang, Yuke Hu, Kemou Li, Jun Liu, Haiwei Wu, Jiantao Zhou, Di Wang·
arXiv:2610.01349v1 Announce Type: cross Abstract: Tool-using large language model (LLM) agents turn generated text into real side effects, so poisoned tool metadata, retrieved pages, memory, and reusable skills can steer the next call. Vetting an artifact before admission does no…
arXiv cs.AI
TIER_1English(EN)·Bingxi Zhao, Lin Geng Foo, Ping Hu, Christian Theobalt, Hossein Rahmani, Jun Liu·
arXiv:2508.17692v2 Announce Type: replace Abstract: Recent advances in LLM-based agents highlight the importance of their reasoning frameworks, which guide the problem-solving process in diverse ways. This survey introduces a unified formal language to systematically categorize t…
arXiv:2605.08386v2 Announce Type: replace Abstract: Skill libraries have become a practical way for LLM agents to reuse procedural experience across tasks. However, existing systems typically treat skills as flat, single-resolution prompt blocks. This creates a tension between re…
arXiv cs.AI
TIER_1English(EN)·Haoyu Wang, Wei Zhao, Yedi Zhang, Christopher M. Poskitt, Jun Sun·
arXiv:2610.00400v1 Announce Type: cross Abstract: Multi-turn attacks on agentic systems can compose individually permissible actions into harmful outcomes, challenging defenses that assess actions or states in isolation. We show that such attacks leave a detectable signature in t…
arXiv:2609.33153v2 Announce Type: replace-cross Abstract: Reported improvements from tools and reusable skills in large language model agents refer to different comparisons. This critical narrative review examines what these evaluations estimate and which conclusions their design…
arXiv cs.AI
TIER_1English(EN)·Bo Deng, Xinlei Zheng, Yi Wei, Kang Zhou, Chongyang Tao, Renzhao Liang, Xuanren Chen, Lifan Guo, Chi Zhang·
arXiv:2610.01256v1 Announce Type: new Abstract: Errors in LLM agent executions and their visible consequences can be separated by many steps, making decisive-error localization a matter of understanding both step content and step dependencies. We introduce DeFA, a dependency-guid…
arXiv cs.AI
TIER_1English(EN)·Haotian Chen, Bowen Ye, Yuning Zhang, Jingkun Yu·
arXiv:2610.01138v1 Announce Type: new Abstract: Concurrent actions in large language model (LLM) agent environments require arbitration even when each proposal is individually valid. We implement a typed snapshot-settlement contract and audit three distinct properties: order sens…
arXiv:2610.01042v1 Announce Type: new Abstract: Multi-agent communication aims to help agents benefit from one another's information. Yet improvements in system performance leave a fundamental ambiguity: do they reflect effective communication, a favorable agent architecture, or …
arXiv:2610.00613v1 Announce Type: new Abstract: Large language model (LLM) based agents are often criticized for lacking spatial understanding and mainly exploiting statistical text patterns. We investigate their spatial comprehension through an architecture combining geometrical…
Multi-agent debate is increasingly used to reach consensus among LLM agents, yet agents often yield to a unanimous majority. When an agent changes its answer, has it changed its mind or only its statement? We study this with two-hop factual questions whose intermediate entity (th…
Multi-agent debate is increasingly used to reach consensus among LLM agents, yet agents often yield to a unanimous majority. When an agent changes its answer, has it changed its mind or only its statement? We study this with two-hop factual questions whose intermediate entity (th…
As LLM agents decide on users' behalf which product to buy, which hotel to book, or which paper to cite, a preference for items from certain sources (the sites or services they come from) shapes what users receive and which sources are selected. We study source preference in end-…
Factories are shifting toward smaller lot sizes with high product customization, requiring frequent re-programming of flexible and reconfigurable automation systems. LLM-based agents can be deployed in two complementary roles: Offline, they generate deterministic production seque…
arXiv:2609.38201v1 Announce Type: new Abstract: Long-running tools can dominate coding-agent latency: compilers, test suites, and repository commands take seconds to minutes while the agent idles. This observation stall presents the same tension that drove out-of-order processors…
arXiv:2605.08693v3 Announce Type: replace Abstract: Skills provide an effective mechanism for improving LLM agents on complex tasks, yet in existing agent frameworks, their creation, refinement, and selection are typically governed by external teachers, hand-designed rules, or au…
arXiv:2609.39149v1 Announce Type: new Abstract: Textual skills enable large language model (LLM) based agents to accumulate reusable procedural knowledge without updating model parameters. Yet existing skill evolution remains largely confined to the text space: an optimizer must …
arXiv cs.CL
TIER_1English(EN)·Deema Alnuhait, Gengyu Wang, Muhammad Khalifa, Hao Peng·
arXiv:2609.39050v1 Announce Type: cross Abstract: As multi-agent systems enter high-stakes domains, the possibility that agents may circumvent safety boundaries is a growing concern. Prior work has examined this risk primarily in adversarial settings, where agents are instructed …
arXiv:2609.38266v1 Announce Type: cross Abstract: Agentic large language models (LLMs) now move money through tools, yet the record of what they did is usually a trace their own process emits beside the effect. Janus puts the record on the effect path. A step's proposal, the verd…
arXiv cs.AI
TIER_1English(EN)·Shuyang Zhang (The Hong Kong Polytechnic University), Jianshuo Chang (The Hong Kong Polytechnic University)·
arXiv:2609.39989v1 Announce Type: new Abstract: Background. A component replacement in a language-model agent changes an execution trajectory, potentially altering later observations, resource use, and recovery opportunities. Different evidence is needed to assess its task-level …
arXiv cs.AI
TIER_1English(EN)·Zuming Zhang, Jie He, Yizhe Zhang, Jeff Z. Pan·
arXiv:2609.39382v1 Announce Type: new Abstract: Textual skills provide reusable guidance for large language model agents, but existing approaches often rely on manually curated skill banks or reinforcement learning with indirect and delayed feedback. We introduce SkillFM (Skill F…
arXiv cs.AI
TIER_1English(EN)·Anmol Kabra, Swathi Saravana Selvam, Albert Gong, Chao Wan, Christian Belardi, Dongyoung Go, Katie Z. Luo, Kilian Q. Weinberger·
arXiv:2609.40221v1 Announce Type: cross Abstract: Training LLM agents with reinforcement learning (RL) is bottlenecked by environments, which must provide verifiable rewards, support long-horizon interaction, and scale cheaply. Existing approaches rely on costly human-curated dat…
arXiv:2609.39065v1 Announce Type: cross Abstract: LLM agents increasingly rely on installable skills, which are packages of instructions, code, and resources that equip them with task-specific capabilities and, once installed, can be automatically invoked across subsequent user t…
Large language models (LLMs) have recently been applied in systems research as a tool to reduce human-intensive engineering effort through cost-efficient automation. Decades of research have produced a rich landscape of concurrency control (CC) protocols, each encoding distinct t…
Before a difficult decision, people often act simply to understand the situation better. We turn an object to see another side, place alternatives next to each other, or change one condition and observe what happens. These actions may not complete the task, but they improve the e…
Background. A component replacement in a language-model agent changes an execution trajectory, potentially altering later observations, resource use, and recovery opportunities. Different evidence is needed to assess its task-level benefit and the contribution of local decision q…
arXiv cs.AI
TIER_1English(EN)·Kehang Zhu, Anand Shah, David Parkes·
arXiv:2609.36365v1 Announce Type: new Abstract: Can interaction formats and textual scaffolds help large language model (LLM) agents make better decisions, and do better decisions come with better explanations? We study these questions in auctions and matching, multi-agent enviro…
arXiv:2609.37470v1 Announce Type: new Abstract: A probability used for a decision should refer to the same event across equivalent requests. We introduce probability contracts, a benchmark connecting exact finite-world posteriors, validated event transformations, and failure-awar…
arXiv:2609.35760v2 Announce Type: replace-cross Abstract: When a large language model (LLM) agent executes the same task, token consumption can vary by over an order of magnitude across runs. The agent chooses its next steps based on tool feedback and intermediate results, while …
arXiv cs.AI
TIER_1English(EN)·Sizhe Chen, Arman Zharmagambetov, David Wagner, Chuan Guo·
arXiv:2507.02735v4 Announce Type: replace-cross Abstract: Prompt injection attacks, where untrusted data contains an injected prompt to manipulate the system, have been listed as the top security threat to AI agents. By fine-tuning on simulated prompt injections, SecAlign, a lead…
arXiv:2609.32965v2 Announce Type: replace Abstract: Multiple agents may often conflict in an organization: for example, one coding agent changes an interface in a repository, but another continues to develop on the old version where existing tests become stale. A conversation can…
arXiv:2605.17734v2 Announce Type: replace Abstract: Equipping LLM agents with reusable skills derived from past experience has become a popular and successful approach for tackling complex and long-horizon tasks. However, such lessons are often encoded as textual guidance that re…
arXiv cs.AI
TIER_1English(EN)·Xiaochuan Li, Ryan Ming, Pranav Setlur, Abhijay Paladugu, Andy Tang, Hao Kang, Shuai Shao, Rong Jin, Chenyan Xiong·
arXiv:2602.18998v2 Announce Type: replace Abstract: LLM agents are increasingly expected to operate as general-purpose systems that resolve real-world user requests, yet their dynamic scaling behavior in realistic environments remains poorly understood. In this paper, we systemat…
arXiv:2609.36739v1 Announce Type: cross Abstract: Multi-agent LLM systems are increasingly structured like organizations, with roles, critics and shared memory, yet they are evaluated on tasks that last minutes. We ask how such an organization behaves when the ground it stands on…
arXiv:2609.36086v1 Announce Type: cross Abstract: Language models are frequently employed to evaluate other language models. An LM evaluator scoring agentic behaviors across multiple criteria is valuable, provided that its decisions align with human judgment. We call the problem …
arXiv cs.AI
TIER_1English(EN)·Lucas Biechy, C\'edric Eichler, H\'eber H. Arcolezi, Nicolas Anciaux·
arXiv:2609.35937v1 Announce Type: cross Abstract: While prior work has documented privacy failures in LLM agents, it remains unclear how the presentation of privacy guidance influences their choice of information sources. We introduce PrivacySkills, a controlled framework for eva…
arXiv:2609.35928v1 Announce Type: cross Abstract: Multi-agent LLM systems increasingly mix models from several providers, yet exposing each agent's underlying model identity to its peers significantly impairs cooperation. We show that when agents are aware of each other's model f…
arXiv:2609.35911v1 Announce Type: cross Abstract: Adapting large language model agents during deployment requires not only retaining past experience, but also turning new observations into timely guidance. Many test-time learning methods, however, acquire knowledge from completed…
arXiv:2609.35813v1 Announce Type: cross Abstract: Compact surrogates could reduce the cost of simulating large language model societies, but must reproduce collective behavior. We compare individual predictions and collective forecasts using 9,455 published trajectories and new e…
arXiv cs.AI
TIER_1English(EN)·Subba Reddy Oota, Francisco Herrera, Jordi Cabot Sagrera, Marcos L\'opez de Prado, Shadab Khan·
arXiv:2609.38108v1 Announce Type: new Abstract: Large language models (LLMs) enable agents to solve long-horizon tasks by generating a plan and then executing it in an environment. However, successful planning requires two distinct capabilities: selecting an appropriate plan for …
arXiv:2609.38043v1 Announce Type: new Abstract: Interactive agent benchmarks and multi-turn reinforcement learning increasingly place a second language model in the role of the user. This simulated user controls what information the agent receives and when, yet current benchmarks…
arXiv:2609.37658v1 Announce Type: new Abstract: LLM agents are increasingly expected to support enterprise workflows, where tasks often involve missing information, uncertainty, feedback, and long-term trade-offs. However, existing enterprise and financial benchmarks mainly test …
arXiv:2609.37356v1 Announce Type: new Abstract: Generating optimization instances that are both feasible and computationally challenging is crucial for benchmarking solvers and training learning-based optimization algorithms. Existing non-LLM generators rely on seed instances or …
arXiv:2609.37172v1 Announce Type: new Abstract: Large language models (LLMs) have emerged as powerful tools for automated heuristic design (AHD), enabling iterative generation and refinement of heuristics. However, the dominant paradigm embeds LLMs as narrow, fixed components, su…
arXiv:2609.37111v1 Announce Type: new Abstract: Long-horizon LLM agents require reinforcement learning methods that can assign credit to intermediate decisions under sparse and delayed rewards. Existing group-based methods such as GRPO and GiGPO alleviate this issue by comparing …
arXiv:2609.36855v1 Announce Type: new Abstract: Multi-agent LLM systems rely on message passing among specialized agents to accomplish complex tasks. However, an upstream agent may provide useful information or an incorrect answer that causes a downstream agent to override a corr…
arXiv:2609.36829v1 Announce Type: new Abstract: An executor can respond strongly to a change in a supplied plan's priority while showing a small change in the same information-selection probability when a default-aligned whole plan is removed. We call the risk of interpreting the…
arXiv:2609.36580v1 Announce Type: new Abstract: LLM agents deployed in real-world environments continually encounter new tasks and safety risks, while execution feedback typically becomes available only after each task is completed. However, existing self-evolving approaches comm…
As multi-agent systems enter high-stakes domains, the possibility that agents may circumvent safety boundaries is a growing concern. Prior work has examined this risk primarily in adversarial settings, where agents are instructed or rewarded to communicate covertly and evade over…
As large language models (LLMs) evolve from standalone models into collaborative agents embedded in physical systems, their reasoning and execution are increasingly distributed across wireless edge nodes. In this setting, wireless networks are experiencing a paradigm shift from o…
Multi-agent LLM systems are increasingly structured like organizations, with roles, critics and shared memory, yet they are evaluated on tasks that last minutes. We ask how such an organization behaves when the ground it stands on keeps moving. Frontier Autolab is a long-horizon …
arXiv:2609.31281v1 Announce Type: new Abstract: Multi-agent cooperative tasks require different agents to execute a joint action simultaneously, and each agent's action affects both the observations and responses of the other agents. Hence, a world model is needed to predict the …
arXiv cs.AI
TIER_1English(EN)·Bovard Doerschuk-Tiberi, Yao Yan, Justin Chiu, Hann Wang, Timothy Chung, Martyna Plomecka, John Schultz, Jon Lipovetz, Clayton Drazner, Yuchen Zhuang, Jaimie Hwang, Nate Keating, Riley Jones, Andrew Lee, Oran Kelly, Ian Gemp, Michael Aaron, Laurel Prince…·
arXiv:2609.31473v1 Announce Type: new Abstract: We introduce Kaggle Game Arena, an open and ever-expanding platform to evaluate large language models (LLMs) through competitive games. Different from static benchmarks, game arena enables models to play head-to-head matchups in str…
arXiv cs.AI
TIER_1English(EN)·Junyi Shen, Noppanat Wadlom, Zhengyuan Su, Yao Lu·
arXiv:2609.31047v1 Announce Type: cross Abstract: Agentic LLM workflows decide their execution paths at runtime. Downstream computation may be predictable, or may have run before, yet it cannot begin until the model or the user resolves the branch. We call this serialization the …
arXiv:2609.31166v1 Announce Type: cross Abstract: Recommender systems have traditionally been developed for platforms. However, this has given rise to many phenomena that may be advantageous for platform lock-in but are a nuisance to users, such as clickbait, filter bubbles, and …
arXiv cs.LG
TIER_1English(EN)·Aparajith Chandran, Juwon Kim, Saurav Jha, Pablo Castells, Florian Hottier·
arXiv:2609.30541v1 Announce Type: new Abstract: Optimizing embedding systems for production recommendation pipelines demands systematic exploration that consumes disproportionate engineering effort at scale. We apply Andrej Karpathy's AutoResearch paradigm -- a large language mod…
arXiv:2604.15186v2 Announce Type: replace-cross Abstract: Agentic workflows carry out complex tasks by orchestrating multiple large language models (LLMs) and tools. Serving them at a target throughput with low latency is hard because they are written in arbitrary agentic framewo…
arXiv cs.AI
TIER_1English(EN)·Jaime Alonso Ruiz, Carlos Aparicio, Gabriel Huecas, Joaqu\'in Salvach\'ua, Andres Munoz-Arcentales·
arXiv:2609.30341v1 Announce Type: new Abstract: Data Spaces enable sovereign and governed data sharing across organizational boundaries, but their integration with AI agents remains challenging due to mismatches between probabilistic language model interactions and policy-driven …
arXiv cs.AI
TIER_1English(EN)·Jinfeng Xu, Zheyu Chen, Ziyue Peng, Zheng Lin, Shuo Yang, Jinze Li, Zheng Xing, Mengran Li, Victor C. M. Leung·
arXiv:2609.30734v1 Announce Type: new Abstract: Multi-agent LLM workflows use planning, execution, verification, and summarization to improve task performance, yet the value of each component depends on the state already produced. Executing every component can waste computation o…
Scientific coding agents receive equations, boundary conditions, and output requirements in writing, then must assess the programs they revise. Rules to Tools (R2T) supplies prepared executable checks of public scientific requirements. Matched SciCode repair groups share written …
Multi-agent LLM systems increasingly mix models from several providers, yet exposing each agent's underlying model identity to its peers significantly impairs cooperation. We show that when agents are aware of each other's model family, the group splits into clusters, where agent…
Long-horizon competition tests agents' ability to coordinate business decisions under uncertainty and adapt to changing rival strategies. We introduce CEO Arena, a benchmark that uses matched replacement evaluation to assess operating returns alongside an agent's effects on rival…
The optimization of LLM serving engines, such as vLLM and SGLang, is largely benchmark-driven: optimizations, scheduling policies, hardware and system designs are all selected based on representative workloads. However, a significant mismatch has emerged in the agentic era. Exist…
Bipartite matching is a fundamental problem in game theory and market design. Classical approaches such as Gale--Shapley assume complete preferences and centralized computation, whereas many real-world matching processes are decentralized, asynchronous, and shaped by sequential i…
LLM-based multi-agent systems (MAS) have shown promise in complex problem solving. As MAS methods diversify, systematic evaluation becomes increasingly challenging. However, existing benchmarks largely focus on final outcomes, leaving unclear how collaboration gains arise, are pr…
Evaluating how LLM agents recover from mid-task failures is central to deploying reliable agentic systems. Existing checkpoint-based benchmarks measure recovery by comparing which action is selected as best across independent runs, a quantity known as set agreement. However, set …
Multi-agent LLM systems increasingly mix models from several providers, yet exposing each agent's underlying model identity to its peers significantly impairs cooperation. We show that when agents are aware of each other's model family, the group splits into clusters, where agent…
Large-language-model multi-agent systems (LLM-MAS) introduce a characteristic reliability problem: an error produced by one agent can be accepted as context by downstream agents and propagate across the workflow. Many proposed safeguards rely on learned or LLM-based judges whose …
Multi-agent systems (MAS) have recently emerged as an effective approach for coordinating large language model (LLM)-based agents to solve complex tasks through structured interactions. In practice, MASs often handle a stream of heterogeneous and complex tasks, requiring agents t…
Large language models (LLMs) are increasingly deployed in multi-agent settings, where agents observe and influence one another, making social influence a key dimension of AI behavior and safety. We investigate whether LLMs' responses depend on the social identity of other agents,…
arXiv cs.MA (Multiagent)
TIER_1English(EN)·Phan Xuan Tan·
Multi-agent LLM systems are increasingly used for deliberation and evaluation, often under the assumption that greater peer interaction leads to more reliable consensus. Existing work largely evaluates these systems through final accuracy or aggregate agreement. However, such mea…
arXiv cs.MA (Multiagent)
TIER_1English(EN)·Christopher G. Brinton·
Multi-agent systems built on large language models (LLMs) are largely homogeneous, as their agents behave alike even across distinct LLMs. We show that when such agents act concurrently without communication, they collide on targets they must split and diverge on targets they mus…
Several works have explored multi-agent reinforcement learning (MARL) in LLM collaboration. However, constructing reliable rewards is difficult in practice, as complete and accurate metrics are often unavailable and hard to aggregate. Preference learning provides an alternative b…
LLM agents often operate over multi-turn interactions in which user intent changes before execution. We study intent drift: the failure mode in which superseded parts of the user's intent continue to influence the final answer or tool action. We introduce IntentFlux, an executabl…
Multiple agents may often conflict in an organization: for example, one coding agent changes an interface in a repository, but another continues to develop on the old version where existing tests become stale. A conversation can resolve the episode, but when the participants chan…
Existing multi-agent benchmarks primarily test in competitive settings, short-horizon interactions under 20 steps, or simply aggregate individual performance, failing to isolate and highlight genuine collaboration capabilities of LLM-based agents. We introduce AgentWorld, a bench…
Recommender systems have traditionally been developed for platforms. However, this has given rise to many phenomena that may be advantageous for platform lock-in but are a nuisance to users, such as clickbait, filter bubbles, and the spread of fake news. Recently, user-side recom…
arXiv:2609.29154v1 Announce Type: new Abstract: Large language model agents increasingly rely on natural-language skills to solve complex tool-use tasks. However, such tasks often admit multiple valid solution paths, making it inappropriate to improve skills by forcing failed tra…
arXiv cs.AI
TIER_1English(EN)·Xingyu Su, Abhishek Kumar, Qing Ping, Youzhi Luo, Jonathan Buck, Zach Zhang, Subramanian Chidambaram, Vinayak Arannil·
arXiv:2609.29051v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) has become a popular recipe for post-training LLM agents. It supervises the agent model at the token level with a stronger teacher view of the same model, obtained by conditioning on privileged inf…
arXiv:2609.29095v1 Announce Type: cross Abstract: When a tool-using agent's write times out or returns a server error, the action may already have taken effect. Retrying blindly duplicates it -- a second charge, a second announcement, a second deployment -- while giving up skips …
arXiv cs.AI
TIER_1English(EN)·Saeedeh Lohrasbi, Mohammad Mamun, Ahmed Yehia, Scott Buffett, Sherif Saad·
arXiv:2609.28559v1 Announce Type: cross Abstract: LLMs increasingly operate through coding-agent harnesses that inspect repositories, invoke tools, and modify files. Substituting the model behind such an agent can therefore change security-relevant decisions, including whether it…
arXiv:2609.29773v1 Announce Type: new Abstract: Many real-world tasks (e.g., office workflows, scientific experimentation) require LLM agents to interact repeatedly with their environments for context-dependent operations. However, such environments are often not agent-ready. Fir…
arXiv cs.AI
TIER_1English(EN)·Igor Bogdanov, Olga Manakina, Chung-Horng Lung·
arXiv:2609.29508v1 Announce Type: new Abstract: Large language model (LLM) agents may perform well on isolated tasks yet drift into inconsistency over extended interaction. We evaluate temporal consistency in a controlled 20-step multi-agent setting inspired by delayed-gratificat…
We introduce Kaggle Game Arena, an open and ever-expanding platform to evaluate large language models (LLMs) through competitive games. Different from static benchmarks, game arena enables models to play head-to-head matchups in structured environments where the gameplay strength…
Existing multi-agent benchmarks primarily test in competitive settings, short-horizon interactions under 20 steps, or simply aggregate individual performance, failing to isolate and highlight genuine collaboration capabilities of LLM-based agents. We introduce AgentWorld, a bench…
Optimizing embedding systems for production recommendation pipelines demands systematic exploration that consumes disproportionate engineering effort at scale. We apply Andrej Karpathy's AutoResearch paradigm -- a large language model that iteratively edits a training script and …
Retrieval-augmented generation (RAG) gives large language models (LLMs) access to external knowledge, but its conventional retrieve-concatenate-generate pipeline makes retrieval decisions on behalf of the model. As tool use and agent loops become more reliable, an agent can decid…
Because significant action to counter global warming requires massive public support, it is important to understand the dynamics of public opinion on climate issues. Of special interest are social tipping points, as revealed by large-scale effects of small perturbations in indivi…
Post-training is becoming a service (PTaaS): a customer hands an operator data and a goal, and a forward-deployed engineer (FDE) returns a fine-tuned, evaluated, and deployed model under a budget, a human-approval gate, and reproducibility requirements. Seating an LLM agent in th…
Every node in a multi-agent large language model (LLM) workflow retrieves context from memory and injects it into its prompt, where those injected tokens are billed as input tokens at the same per-token price as the system prompt and the user query. Production observability tools…
arXiv cs.MA (Multiagent)
TIER_1English(EN)·Gregory B. Rehm·
Multi-agent LLM systems are increasingly evaluated in social dilemmas, but most work treats governance as imposed by the experimenter, expressed rhetorically, or restricted to a fixed menu of mechanisms. We introduce GovSim-SelfGovern, an extension of the GovSim common-pool resou…
Hacker News — AI stories ≥50 points
TIER_1English(EN)·nisosguy·
<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Av0vtrz3qR8UBVFY7rTtjw.png" /><figcaption>Source: Image by the author</figcaption></figure><p>Imagine giving an AI model access to your codebase and asking it to fix a production bug.</p><p>The model understands …
<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sbIZ3OgCJXTZWvVeJDyYbA.jpeg" /></figure><h4><strong>The Black Box Problem Just Got Bigger</strong></h4><p>You deployed your LLM application to production. Users are interacting with it. Tokens are burning. And so…
<p>Tuần này trên Hacker News có một bài khá hot: DeepSeek ra API search và fetch với giá $0.75 cho 1000 request thành công. Đọc comment thì thấy nhiều người tính nhẩm ngay: agent của mình mỗi câu hỏi gọi 5-10 lần search, mỗi lần lại fetch 3-4 trang, vậy một ngày tốn bao nhiêu? Cù…
<p>There's a big difference between an AI that tells you what to do and an AI that actually does it.</p> <p>If a chatbot gives you the wrong answer, you might lose a few minutes correcting it. But what happens when an AI agent misunderstands your request and deletes a customer re…
<p>If you self-host large language models, you have probably settled into one of two camps. You use vLLM or SGLang because they are easy to run, or you use TensorRT-LLM because it squeezes the most out of NVIDIA hardware and you are willing to pay for that in setup pain.</p> <p>T…
<p>Documentation, READMEs, and API references have long helped developers get started with new tools. As coding agents take on more of the work, new formats are emerging to help them navigate that information and use developer products successfully.</p> <p>You've probably encount…
<p>When you're building an LLM agent, especially for complex real-world tasks like processing claims or generating detailed reports, you often hit a wall. A single, monolithic prompt asking the agent to "do X, Y, and Z" usually leads to inconsistent outputs, missed steps, or outr…
<p>Retell AI prices the brain of your voice agent by the minute: the LLM line on its pricing page runs from $0.045 to $0.16 per minute for frontier models, on top of the voice engine. Retell also has a custom LLM slot that skips that line entirely: it opens a WebSocket to your se…
<p>Imagine a coding agent that begins migrating a service on Monday. It identifies an incompatible dependency, documents an alternative approach, and parks the task overnight. When the agent resumes on Tuesday, it forgets the decision entirely, repeats the investigation, and atte…
dev.to — LLM tag
TIER_1English(EN)·Prabod Rathnayaka·
<p>The future of local AI is agentic and requires multiple agents to share one machine: one agent summarising your emails overnight, one watching your health data and one doing the shopping, while you are still able to chat and ask questions in between. Today's inference servers …
dev.to — LLM tag
TIER_1Português(PT)·Silviu Technology·
<p>Los agentes LLM que prueban un registro o un flujo de recuperación de contraseña tratan el inbox temporal como si fuera una caja negra: crean una dirección, preguntan si llegó algo y repiten hasta que aparece un mensaje. Ese patrón parece sencillo, pero mezcla adquisición, esp…
<h2> A quick test </h2> <p>You have two McDonald's coupons and can use only one:</p> <ul> <li>Spend 40, get 10 off</li> <li>12% off any burger</li> </ul> <p>You order a ¥25 burger:</p> <ul> <li>Threshold coupon: you'd have to add ¥15 more → pay ¥30</li> <li>Percentage coupon: 25 …
<p>Most agent frameworks make the LLM the bus between every tool call. Ask an agent to read every Markdown file and extract headings, and the default execution path looks like this: one inference to find the files, another to read file A, another to read file B, another to filter…
<h1> AX-RAY: VIDRAFT's Agent Safety Benchmark Flags 92% of Tested LLMs as Dangerous in Agentic Contexts </h1> <blockquote> <p><strong>TL;DR:</strong> VIDRAFT, a Korean Pre-AGI AI startup based at Seoul AI Hub, has published results from its AI safety diagnostic platform <strong>A…
<p><strong>I turned Andrej Karpathy's tips on understanding LLM output into an Open Source Agent Skill : <code>[make-it-click](https://github.com/ajithraghavan/make-it-click)</code> 🧠⚡</strong></p> <p>Inspired by his X post : as Models do more of the work, we spend more time <em>…
dev.to — LLM tag
TIER_1Español(ES)·Silviu Technology·
<p>Un agente LLM que trabaja con email puede fallar de varias maneras antes de que el evaluador lo note. Puede consultar demasiado pronto, repetir una llamada que ya tuvo éxito o leer un mensaje correcto de una ejecución anterior. Al final vemos un resultado rojo, pero no sabemos…
<p>Your LLM-powered app is in production. Users are hitting it. Something is slow — or wrong — and you have no idea what. You can't reproduce it locally, and your generic APM shows... a single HTTP call to an external API. That's the problem with LLM observability today: the gap …
dev.to — LLM tag
TIER_1Français(FR)·Silviu Technology·
<p>Un agente LLM puede completar un registro, pedir un código de verificación y afirmar que terminó correctamente. Eso no demuestra que el flujo haya funcionado. Quizá leyó un mensaje viejo, confundió una respuesta del sistema o inventó el resultado después de que una herramienta…
dev.to — LLM tag
TIER_1Español(ES)·Silviu Technology·
<p>Los agentes LLM pueden ayudar a investigar por qué falla una prueba de email, pero hay una diferencia importante entre asistir una ejecución y controlar una ejecución. Si el modelo puede hacer cualquier cosa, el resultado será dificil de reproducir: una corrida pasa, la siguie…
<p>Hi, I'm Dr. Haina Fatima. I started my career as a physician and radiologist, and today I lead development and product at XINI8 Engine, an AI-native software engineering platform within the XINI8 ecosystem. Coming from outside traditional software, the terms "LLM" and "AI agen…
dev.to — LLM tag
TIER_1English(EN)·Prabhakar Chaudhary·
<h1> Beyond One-Shot Success: How AREX-2 Teaches LLM Agents to Reflect and Persevere </h1> <p>Current autonomous LLM agents are often evaluated by their ability to solve a task in a single pass or through a short sequence of scripted interactions. While models like GPT-4o and Cla…
dev.to — LLM tag
TIER_1Español(ES)·Silviu Technology·
<p>Un agente LLM puede parecer muy capaz durante una demostración y volverse dificil de operar cuando empieza a llamar APIs, crear tickets o consultar datos reales. El problema no suele ser que el modelo “no sepa” la respuesta. Es que el sistema no define con precisión qué puede …
dev.to — LLM tag
TIER_1English(EN)·Quoc Bao An Nguyen·
<h2> 1. Introduction </h2> <p>When I first started building my AI-powered course recommendation system, I thought integrating an LLM with backend APIs would be straightforward.</p> <p>However, I quickly ran into a key problem:<br /> The model was calling backend APIs almost every…
Fresh on Hacker News (130+ pts, 72 comments): Hybrid Reasoning Agents — LLM agents that reason in latent space and act through text. Today's models 'think out loud': every CoT step becomes tokens — transparent but slow and costly. The twist: keep intermediate reasoning as vectors…
Abstract LLM-driven agent-based models (ABMs) replace a closed-form decision rule with an LLM agent. The idea is that the agent reproduces the qualitative behaviour of the original rule while adding context-sensitive reasoning. We test that assumption on Epstein's (2002) civil vi…
Obelisk 0.42 introduit des agents persistants et des sandboxes en couches pour les LLM. L'angle sécurité est concret : isoler les capacités des agents IA, c'est réduire la surface d'attaque quand un modèle est manipulé ou produit du code inattendu. Le sandboxing multicouche, vieu…
<table> <tr><td> <a href="https://www.reddit.com/r/OpenAI/comments/1x3ezj7/1532_agent_tasks_9_frontier_llms_induce_the_model/"> <img alt="1,532 agent tasks, 9 frontier LLMs: induce the model and its deception rate rises across every task family" src="https://external-preview.redd…
