Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2507.21017 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916867396861952 |
|---|---|
| author | Zhang, Weichen Sun, Yiyou Huang, Pohao Pu, Jiayue Lin, Heyue Song, Dawn |
| author_facet | Zhang, Weichen Sun, Yiyou Huang, Pohao Pu, Jiayue Lin, Heyue Song, Dawn |
| contents | Hallucinations pose critical risks for large language model (LLM)-based agents, often manifesting as hallucinative actions resulting from fabricated or misinterpreted information within the cognitive context. While recent studies have exposed such failures, existing evaluations remain fragmented and lack a principled testbed. In this paper, we present MIRAGE-Bench--Measuring Illusions in Risky AGEnt settings--the first unified benchmark for eliciting and evaluating hallucinations in interactive LLM-agent scenarios. We begin by introducing a three-part taxonomy to address agentic hallucinations: actions that are unfaithful to (i) task instructions, (ii) execution history, or (iii) environment observations. To analyze, we first elicit such failures by performing a systematic audit of existing agent benchmarks, then synthesize test cases using a snapshot strategy that isolates decision points in deterministic and reproducible manners. To evaluate hallucination behaviors, we adopt a fine-grained-level LLM-as-a-Judge paradigm with tailored risk-aware prompts, enabling scalable, high-fidelity assessment of agent actions without enumerating full action spaces. MIRAGE-Bench provides actionable insights on failure modes of LLM agents and lays the groundwork for principled progress in mitigating hallucinations in interactive environments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_21017 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MIRAGE-Bench: LLM Agent is Hallucinating and Where to Find Them Zhang, Weichen Sun, Yiyou Huang, Pohao Pu, Jiayue Lin, Heyue Song, Dawn Artificial Intelligence Hallucinations pose critical risks for large language model (LLM)-based agents, often manifesting as hallucinative actions resulting from fabricated or misinterpreted information within the cognitive context. While recent studies have exposed such failures, existing evaluations remain fragmented and lack a principled testbed. In this paper, we present MIRAGE-Bench--Measuring Illusions in Risky AGEnt settings--the first unified benchmark for eliciting and evaluating hallucinations in interactive LLM-agent scenarios. We begin by introducing a three-part taxonomy to address agentic hallucinations: actions that are unfaithful to (i) task instructions, (ii) execution history, or (iii) environment observations. To analyze, we first elicit such failures by performing a systematic audit of existing agent benchmarks, then synthesize test cases using a snapshot strategy that isolates decision points in deterministic and reproducible manners. To evaluate hallucination behaviors, we adopt a fine-grained-level LLM-as-a-Judge paradigm with tailored risk-aware prompts, enabling scalable, high-fidelity assessment of agent actions without enumerating full action spaces. MIRAGE-Bench provides actionable insights on failure modes of LLM agents and lays the groundwork for principled progress in mitigating hallucinations in interactive environments. |
| title | MIRAGE-Bench: LLM Agent is Hallucinating and Where to Find Them |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2507.21017 |