{"ID":6536205,"CreatedAt":"2026-07-14T01:21:01.169441415Z","UpdatedAt":"2026-07-15T03:28:55.185153975Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.10738","arxiv_id":"2607.10738","title":"To Answer or to Abstain: Mitigating Search-Agent Hallucinations via Abstention-Aware Reinforcement Learning","abstract":"Recent advances in equipping Large Language Models (LLMs) with search tools and outcome-reward reinforcement learning (RL) have achieved new state-of-the-art results on open-domain QA tasks. However, we argue that current training paradigms harbor a critical vulnerability: they predominantly reward correct answers but fail to penalize fabricated ones when retrieval fails, thereby implicitly exacerbating hallucinations. To address this, we propose Abstention-Aware Reinforcement Learning (AWA-RL), which dynamically shapes the abstention reward utilizing the model's query-specific prior capabilities and continuous on-policy training observations. We also introduce a novel metric, RA-F1, to measure the capability-reliability trade-off. Compared to non-abstaining baselines, AWA-RL boosts absolute precision by up to 10.3% and overall RA-F1 by 2.9%, with only marginal sacrifice in raw accuracy. These results confirm that AWA-RL successfully yields highly capable and reliable search agents. The code, data, and model weights are publicly available at https://github.com/zfj1998/AWA-RL.","short_abstract":"Recent advances in equipping Large Language Models (LLMs) with search tools and outcome-reward reinforcement learning (RL) have achieved new state-of-the-art results on open-domain QA tasks. However, we argue that current training paradigms harbor a critical vulnerability: they predominantly reward correct answers but...","url_abs":"https://arxiv.org/abs/2607.10738","url_pdf":"https://arxiv.org/pdf/2607.10738v1","authors":"[\"Fengji Zhang\",\"Tianyu Fan\",\"Yuxiang Zheng\",\"Xinyao Niu\",\"Chengen Huang\",\"Jacky Keung\",\"Bei Chen\"]","published":"2026-07-12T12:43:12Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.CL\"]","methods":"[\"Reinforcement Learning\",\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":614147,"CreatedAt":"2026-07-14T01:21:01.169441415Z","UpdatedAt":"2026-07-14T01:21:01.169441415Z","DeletedAt":null,"paper_id":6536205,"paper_url":"https://arxiv.org/abs/2607.10738","paper_title":"To Answer or to Abstain: Mitigating Search-Agent Hallucinations via Abstention-Aware Reinforcement Learning","repo_url":"https://github.com/zfj1998/AWA-RL","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}