Search Agents
PulseAugur coverage of Search Agents — every cluster mentioning Search Agents across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
SearchJev model boosts search agent speed and accuracy
Researchers have developed SearchJev, a novel System-1 model designed to enhance the efficiency and reliability of search agents. This model separates quick decision-making from complex reasoning, scoring potential sear…
-
New research proposes Confluent Evidence Acquisition for AI agents
A new research paper introduces Confluent Evidence Acquisition (CEA), a method designed to improve how language-model agents gather information during complex tasks. Current agents can be overly influenced by the timing…
-
New RL methods enhance LLM reasoning with fine-grained credit assignment · 9 sources tracked
Researchers are developing new methods for credit assignment in reinforcement learning (RL) for large language models (LLMs) to improve their reasoning and decision-making capabilities. Several papers propose novel fram…
-
New framework improves LLM search agent training with intermediate rewards
Researchers have developed a new training framework to improve the efficiency of search agents used by large language models (LLMs). The framework addresses the challenge of credit assignment in reinforcement learning b…
-
New method improves reinforcement learning for search agents
Researchers have developed a new method for training reinforcement learning agents used in search tasks. This approach, called dense process supervision via fact utility estimation, addresses the challenge of assigning …
-
New benchmark DiscoBench evaluates LLM search agents' clarification skills
A new benchmark called DiscoBench has been introduced to evaluate the ability of search agents, particularly those powered by LLMs, to handle ambiguous queries. DiscoBench assesses how effectively these agents can ident…
-
Google rolls out AI-powered Search agents for information tracking
Google is beginning to roll out "Search agents" within its AI Mode, specifically for AI Ultra subscribers. These agents are designed to proactively track and manage information for users. The concept was initially revea…
-
New Benchmark Evaluates Search Agent Performance on Daily Tasks
Researchers have developed DailyReport, a new benchmark designed to evaluate the capabilities of search agents (SAs) on realistic, open-ended daily search tasks. Unlike previous benchmarks that focused on specialized sc…
-
New DailyReport benchmark evaluates search agent performance on real-world tasks
Researchers have introduced DailyReport, a new open-ended benchmark designed to evaluate the capabilities of search agents on everyday information-seeking tasks. This benchmark includes 150 tasks and 3,546 rubrics, aimi…
-
New CalibAdv Method Enhances Search Agent Training Stability
A new method called CalibAdv has been developed to improve the training stability and performance of search agents, particularly those using Group Relative Policy Optimization (GRPO). This approach addresses issues wher…