analiza aiqbench
This research introduces Cross-Layer Attention Probing (CLAP), a novel technique for detecting hallucinations in Large Language Models by analyzing activations across the entire residual stream as a joint sequence. Empirical results demonstrate that CLAP enables fine-grained disambiguation of hallucinated versus non-hallucinated responses, facilitating a detect-then-mitigate strategy to improve model reliability even under out-of-distribution conditions.
Cross-Layer Attention Probing for Fine-Grained Hallucination Detection
arXiv:2509.09700v2 Announce Type: replace Abstract: With the large-scale adoption of Large Language Models (LLMs) in various applications, there is a growing reliability concern due to their tendency to generate inaccurate text, i.e. hallucinations. In this work, we propose Cross-Layer Attention Probing (CLAP), a novel activation probing technique for hallucination detection, which processes the LLM activations across the entire residual stream as a joint sequence. Our empirical evaluations using five LLMs and three tasks show that CLAP improves hallucination detection compared to baselines on both greedy decoded responses as well as responses sampled at higher temperatures, thus enabling fine-grained detection, i.e. the ability to disambiguate hallucinations and non-hallucinations among different sampled responses to a given prompt. This allows us to propose a detect-then-mitigate strategy using CLAP to reduce hallucinations and improve LLM reliability compared to direct mitigation approaches. Finally, we show that CLAP maintains high reliability even when applied out-of-distribution. Code available at: https://github.com/itsmemala/CLAP
Abstract:With the large-scale adoption of Large Language Models (LLMs) in various applications, there is a growing reliability concern due to their tendency to generate inaccurate text, i.e. hallucinations. In this work, we propose Cross-Layer Attention Probing (CLAP), a novel activation probing technique for hallucination detection, which processes the LLM activations across the entire residual stream as a joint sequence. Our empirical evaluations using five LLMs and three tasks show that CLAP improves hallucination detection compared to baselines on both greedy decoded responses as well as responses sampled at higher temperatures, thus enabling fine-grained detection, i.e. the ability to disambiguate hallucinations and non-hallucinations among different sampled responses to a given prompt. This allows us to propose a detect-then-mitigate strategy using CLAP to reduce hallucinations and improve LLM reliability compared to direct mitigation approaches. Finally, we show that CLAP maintains high reliability even when applied out-of-distribution. Code available at: this https URL
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2509.09700 [cs.CL]
(or arXiv:2509.09700v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2509.09700 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 246 cuvinte
Subiecte conexe
External Observers May See More Clearly: Cross-Model Span-Level Hallucination Detection in Large Language Models via Hidden State Probing
arXiv · 2 oct. 2026
Analyzing the Correlation Between Hallucinations and Knowledge Conflicts in Large Language Models
arXiv · 9 iun. 2026
AgentOps: Operationalize agentic AI at scale with Amazon Bedrock AgentCore
AWS · 1 iun. 2026
Some Intuitions on Steering Vectors
LessWrong · 29 sept. 2026