Can a Small Model Learn to Look Before It Leaps? Dynamic Learning and Proactive Correction for Hallucination Detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915832658919424 |
|---|---|
| author | Bao, Zepeng Zhou, Shen Pi, Qiankun Chen, Jianhao Xu, Mayi Zhong, Ming Zhu, Yuanyuan Qian, Tieyun |
| author_facet | Bao, Zepeng Zhou, Shen Pi, Qiankun Chen, Jianhao Xu, Mayi Zhong, Ming Zhu, Yuanyuan Qian, Tieyun |
| contents | Hallucination in large language models (LLMs) remains a critical barrier to their safe deployment. For hallucination detection to be practical in real-world scenarios, the use of efficient small models is essential to ensure low latency and minimal resource consumption. However, existing methods rely on fixed verification strategies, where simply tuning small models to mimic fixed verification trajectories fails to capture the adaptability required for diverse hallucination patterns, thereby inducing planning instability. To address this limitation, we propose a ``Learning to Evaluate and Adaptively Plan'' (LEAP) framework, which shifts hallucination detection from fixed execution to dynamic strategy learning. Specifically, LEAP first employs a powerful teacher model to iteratively explore and refine verification strategies through a failure-driven loop. This dynamic planning capability is then distilled into an efficient student model, augmented by a novel proactive correction mechanism that enables the model to evaluate and optimize its verification strategy before execution. Experiments on three benchmarks demonstrate that LEAP outperforms state-of-the-art methods, offering an effective and scalable solution for reliable hallucination detection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_05854 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Can a Small Model Learn to Look Before It Leaps? Dynamic Learning and Proactive Correction for Hallucination Detection Bao, Zepeng Zhou, Shen Pi, Qiankun Chen, Jianhao Xu, Mayi Zhong, Ming Zhu, Yuanyuan Qian, Tieyun Artificial Intelligence Hallucination in large language models (LLMs) remains a critical barrier to their safe deployment. For hallucination detection to be practical in real-world scenarios, the use of efficient small models is essential to ensure low latency and minimal resource consumption. However, existing methods rely on fixed verification strategies, where simply tuning small models to mimic fixed verification trajectories fails to capture the adaptability required for diverse hallucination patterns, thereby inducing planning instability. To address this limitation, we propose a ``Learning to Evaluate and Adaptively Plan'' (LEAP) framework, which shifts hallucination detection from fixed execution to dynamic strategy learning. Specifically, LEAP first employs a powerful teacher model to iteratively explore and refine verification strategies through a failure-driven loop. This dynamic planning capability is then distilled into an efficient student model, augmented by a novel proactive correction mechanism that enables the model to evaluate and optimize its verification strategy before execution. Experiments on three benchmarks demonstrate that LEAP outperforms state-of-the-art methods, offering an effective and scalable solution for reliable hallucination detection. |
| title | Can a Small Model Learn to Look Before It Leaps? Dynamic Learning and Proactive Correction for Hallucination Detection |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2511.05854 |