New AI research tackles LLM hallucinations with novel detection and intervention methods
ByPulseAugur Editorial·
Summary by gemini-2.5-flash-lite
from 53 sources
Researchers are developing novel methods to combat hallucinations in Large Language Models (LLMs). Several papers propose new frameworks and techniques, including LaaB, which bridges neural features and symbolic judgments, and CuraView, a multi-agent system for medical hallucination detection using GraphRAG. Other approaches focus on neuro-symbolic agents for hallucination-free requirements reuse, adaptive unlearning for surgical hallucination suppression in code generation, and harnessing reasoning trajectories via answer-agreement representation shaping. Additionally, new benchmarks like HalluScan are being created to systematically evaluate detection and mitigation strategies.
AI
IMPACT
New research offers diverse strategies to improve LLM factual accuracy, crucial for reliable deployment in sensitive domains like healthcare and code generation.
RANK_REASON
Multiple arXiv papers present novel research on detecting and mitigating LLM hallucinations.
Large language models hallucinate during multi-step reasoning, but most existing detectors operate at the trace level: they assign one confidence score to a full output, fail to localize the first error, and often require multiple sampled completions. We frame hallucination inste…
With the vast amount of content uploaded every hour, along with the AI generated content that can include hallucinations, Automated Fact-Checking (AFC) has become increasingly vital, as it is infeasible for human fact-checkers to manually verify the sheer volume of information ge…
Large language models (LLMs) have demonstrated remarkable capabilities, but they still frequently produce hallucinations. These hallucinations are difficult to detect in reasoning-intensive tasks, where the content appears coherent but contains errors like logical flaws and unrel…
Multimodal large language models (MLLMs) have achieved remarkable progress, yet the object hallucination remains a critical challenge for reliable deployment. In this paper, we present an in-depth analysis of instruction token embeddings and reveal that they implicitly encode vis…
Large language models (LLMs) are known to generate plausible but false information across a wide range of contexts, yet the real-world magnitude and consequences of this hallucination problem remain poorly understood. Here we leverage a uniquely verifiable object - scientific cit…
arXiv cs.CL
TIER_1·Brandon C. Colelough, Davis Bartels, Dina Demner-Fushman·
arXiv:2603.09986v2 Announce Type: replace Abstract: Hallucinations, the tendency for large language models to provide responses with factually incorrect and unsupported claims, is a serious problem within natural language processing for which we do not yet have an effective solut…
arXiv cs.CL
TIER_1·Erik Nielsen, Elia Cunegatti, Marcus Vukojevic, Giovanni Iacca·
arXiv:2605.05953v1 Announce Type: new Abstract: One of the most critical challenges in Large Language Models is their tendency to hallucinate, i.e., produce factually incorrect responses. Existing approaches show promising results in terms of hallucination correction, but still s…
One of the most critical challenges in Large Language Models is their tendency to hallucinate, i.e., produce factually incorrect responses. Existing approaches show promising results in terms of hallucination correction, but still suffer from a main limitation: they apply correct…
One of the most critical challenges in Large Language Models is their tendency to hallucinate, i.e., produce factually incorrect responses. Existing approaches show promising results in terms of hallucination correction, but still suffer from a main limitation: they apply correct…
arXiv cs.LG
TIER_1·Linggang Kong, Lei Wu, Yunlong Zhang, Xiaofeng Zhong, Zhen Wang, Yongjie Wang, Yao Pan·
arXiv:2604.11087v2 Announce Type: replace Abstract: Despite the groundbreaking advancements made by large language models (LLMs), hallucination remains a critical bottleneck for their deployment in high-stakes domains. Existing classification-based methods mainly rely on static a…
arXiv:2605.05166v1 Announce Type: new Abstract: Self-consistency detects hallucinations by generating multiple sampled answers to a question and measuring agreement, but this requires repeated decoding and can be sensitive to lexical variation. Semantic self-consistency improves …
arXiv:2605.05025v1 Announce Type: new Abstract: We propose a lightweight and single-pass uncertainty quantification method for detecting hallucinations in Large Language Models. The method uses attention matrices to estimate uncertainty without requiring repeated sampling or exte…
arXiv:2605.05134v1 Announce Type: new Abstract: Large Language Models (LLMs) frequently generate plausible but non-factual content, a phenomenon known as hallucination. While existing detection methods typically rely on computationally expensive sampling-based consistency checks …
arXiv:2605.05045v1 Announce Type: cross Abstract: Vision-language models (VLMs) achieve strong multimodal performance but remain prone to relation hallucination, which requires accurate reasoning over inter-object interactions. We study the impact of visual perturbations, specifi…
Self-consistency detects hallucinations by generating multiple sampled answers to a question and measuring agreement, but this requires repeated decoding and can be sensitive to lexical variation. Semantic self-consistency improves this by clustering sampled answers by meaning us…
Large Language Models (LLMs) frequently generate plausible but non-factual content, a phenomenon known as hallucination. While existing detection methods typically rely on computationally expensive sampling-based consistency checks or external knowledge retrieval, we propose a ne…
We propose a lightweight and single-pass uncertainty quantification method for detecting hallucinations in Large Language Models. The method uses attention matrices to estimate uncertainty without requiring repeated sampling or external models. Specifically, we measure the Kullba…
arXiv:2605.01562v1 Announce Type: cross Abstract: The Object-Oriented Method for Requirements Authoring and Management (OOMRAM) is a requirements reuse framework that relies on exact identifier matching and rigid templates, limiting its ability to adapt specifications across dive…
arXiv:2605.03971v1 Announce Type: new Abstract: Large Language Models (LLMs) are prone to factual hallucinations, risking their reliability in real-world applications. Existing hallucination detectors mainly extract micro-level intrinsic patterns for uncertainty quantification or…
arXiv:2605.03476v1 Announce Type: new Abstract: Discharge summaries require extracting critical information from lengthy electronic health records (EHRs), a process that is labor-intensive when performed manually. Large language models (LLMs) can improve generation efficiency; ho…
Large Language Models (LLMs) are prone to factual hallucinations, risking their reliability in real-world applications. Existing hallucination detectors mainly extract micro-level intrinsic patterns for uncertainty quantification or elicit macro-level self-judgments through verba…
Discharge summaries require extracting critical information from lengthy electronic health records (EHRs), a process that is labor-intensive when performed manually. Large language models (LLMs) can improve generation efficiency; however, they are prone to producing faithfulness …
Discharge summaries require extracting critical information from lengthy electronic health records (EHRs), a process that is labor-intensive when performed manually. Large language models (LLMs) can improve generation efficiency; however, they are prone to producing faithfulness …
arXiv cs.LG
TIER_1·Yee Zhing Liew, Andrew Huey Ping Tan, Anwar P. P Abdul Majeed·
arXiv:2605.00939v1 Announce Type: new Abstract: Traditional hallucination detection fails on "Stubborn Hallucinations" -- errors where LLMs are confidently wrong. We propose a geometric solution: Embedding-Perturbed Gradient Sensitivity (EPGS). We hypothesize that while robust fa…
arXiv cs.LG
TIER_1·Jianxiong Zhang, Bing Guo, Yuming Jiang, Haobo Wang, Bo An, Sean Du·
arXiv:2605.02443v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse natural language processing tasks, yet they remain susceptible to hallucinations -- generating content that is factually incorrect, unfaithful to …
arXiv cs.CL
TIER_1·Freja Thoresen, Dan Saattrup Smart·
arXiv:2605.02504v1 Announce Type: new Abstract: Most hallucination evaluations focus on English, leaving it unclear whether findings transfer to lower-resource languages. We investigate faithfulness hallucinations, defined as model-generated content that is fluent and plausible b…
arXiv cs.CL
TIER_1·Alexandra Bazarova, Aleksandr Yugay, Andrey Shulga, Alina Ermilova, Andrei Volodichev, Konstantin Polev, Julia Belikova, Rauf Parchiev, Dmitry Simakov, Maxim Savchenko, Andrey Savchenko, Serguei Barannikov, Alexey Zaytsev·
arXiv:2504.10063v4 Announce Type: replace Abstract: Hallucination, i.e., generating factually incorrect content, remains a critical challenge for large language models (LLMs). We introduce TOHA, a TOpology-based HAllucination detector in the RAG setting, which leverages a topolog…
arXiv:2605.01047v1 Announce Type: cross Abstract: Hallucinations, outputs that sound plausible but are factually incorrect, remain an open challenge for deployed LLMs. In code generation, models frequently hallucinate non-existent software packages, recommending imports and insta…
Most hallucination evaluations focus on English, leaving it unclear whether findings transfer to lower-resource languages. We investigate faithfulness hallucinations, defined as model-generated content that is fluent and plausible but diverges from the provided input or is intern…
Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse natural language processing tasks, yet they remain susceptible to hallucinations -- generating content that is factually incorrect, unfaithful to provided context, or misaligned with user instru…
Multimodal large language models (MLLMs) have revolutionized the landscape of AI, demonstrating impressive capabilities in tackling complex vision and audio-language tasks. However, a critical challenge remains: these models often suffer from hallucinations, generating outputs th…
arXiv cs.CL
TIER_1·Guoshenghui Zhao, Weijie Zhao, Tan Yu·
arXiv:2604.26139v1 Announce Type: new Abstract: Diffusion large language models generate text through multi-step denoising, where hallucination signals may emerge throughout the trajectory rather than only in the final output. Existing detectors mainly rely on output uncertainty …
arXiv cs.CL
TIER_1·Jiawei Li, Akshayaa Magesh, Venugopal V. Veeravalli·
arXiv:2508.18473v3 Announce Type: replace Abstract: While Large Language Models (LLMs) have emerged as powerful foundational models to solve a variety of tasks, they have also been shown to be prone to hallucinations, i.e., generating responses that sound confident but are actual…
Diffusion large language models generate text through multi-step denoising, where hallucination signals may emerge throughout the trajectory rather than only in the final output. Existing detectors mainly rely on output uncertainty or coarse trace statistics, which often fail to …
arXiv:2604.23366v1 Announce Type: new Abstract: Autonomous multi-agent LLM systems are increasingly deployed to investigate operational incidents and produce structured diagnostic reports. Their trustworthiness hinges on whether each claim is grounded in observed evidence rather …
Vision-Language Models (VLMs) are frequently undermined by object hallucination--generating content that contradicts visual reality--due to an over-reliance on linguistic priors. We introduce Positive-and-Negative Decoding (PND), a training-free inference framework that intervene…
Autonomous multi-agent LLM systems are increasingly deployed to investigate operational incidents and produce structured diagnostic reports. Their trustworthiness hinges on whether each claim is grounded in observed evidence rather than model-internal inference. Existing grounded…
Vision-language models (VLMs) have demonstrated remarkable capabilities in bridging visual perception and natural language understanding, enabling a wide range of multimodal reasoning tasks. However, they often produce object hallucinations, describing content absent from the inp…
Large Vision-Language Models (LVLMs) have achieved remarkable progress in multimodal tasks, yet their reliability is persistently undermined by hallucinations-generating text that contradicts visual input. Recent studies often attribute these errors to inadequate visual attention…
arXiv:2603.04673v2 Announce Type: replace-cross Abstract: Deep learning (DL) methods are currently being explored to restore images from sparse-view-, limited-data-, and undersampled-based acquisitions in medical applications. Although outputs from DL may appear visually appealin…
Vision-language models (VLMs) achieve strong multimodal performance but remain prone to relation hallucination, which requires accurate reasoning over inter-object interactions. We study the impact of visual perturbations, specifically rotation and noise, and show that even mild …
arXiv:2605.01766v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) have revolutionized the landscape of AI, demonstrating impressive capabilities in tackling complex vision and audio-language tasks. However, a critical challenge remains: these models often…
arXiv:2511.20032v3 Announce Type: replace Abstract: Visual attention serves as the primary mechanism through which MLLMs interpret visual information; however, its limited localization capability often leads to hallucinations. We observe that although MLLMs can accurately extract…
arXiv:2604.25642v1 Announce Type: new Abstract: Large Vision-Language Models (LVLMs) have achieved remarkable progress in visual-textual understanding, yet their reliability is critically undermined by hallucinations, i.e., the generation of factually incorrect or inconsistent re…
Large Vision-Language Models (LVLMs) have achieved remarkable progress in visual-textual understanding, yet their reliability is critically undermined by hallucinations, i.e., the generation of factually incorrect or inconsistent responses. While recent studies using steering vec…
arXiv:2604.24396v1 Announce Type: new Abstract: Vision-Language Models (VLMs) are frequently undermined by object hallucination--generating content that contradicts visual reality--due to an over-reliance on linguistic priors. We introduce Positive-and-Negative Decoding (PND), a …
arXiv:2406.10185v2 Announce Type: replace Abstract: Large Vision Language Models (LVLMs) are increasingly integral to healthcare applications, including medical visual question answering and imaging report generation. While these models inherit the robust capabilities of foundati…
arXiv:2604.22822v1 Announce Type: new Abstract: Object level hallucination remains a central reliability challenge for vision language models (VLMs), particularly in binary object existence verification. Existing benchmarks emphasize aggregate accuracy but rarely disentangle whet…
Vision-Language Models (VLMs) are frequently undermined by object hallucination--generating content that contradicts visual reality--due to an over-reliance on linguistic priors. We introduce Positive-and-Negative Decoding (PND), a training-free inference framework that intervene…
<p> </p> <p>E aí, gurizada! Tudo tranquilo? Hoje eu quero trocar uma ideia com vocês sobre umas paradas que andei percebendo com as IAs, e que me motivaram a gravar um vídeo e até escrever um post lá no meu site, o marcelocabral.com.br. Sabe quando a inteligência artificial solta…