QueryBandits 用于幻觉缓解：利用语义特征实现无遗憾重写
QueryBandits for Hallucination Mitigation: Exploiting Semantic Features for No-Regret Rewriting
AI总结：
本文提出 QueryBandits 赌博机框架，利用查询的 17 个语义特征动态选择重写策略以主动缓解大语言模型幻觉，在 13 个问答基准上显著优于无重写基线和静态提示方法，并证明静态重写可能加剧幻觉且不存在全局最优的单一重写策略。
摘要：
大语言模型（LLMs）中高级推理能力的提升导致了更高的幻觉发生率；然而大多数缓解工作侧重于事后过滤，而不是塑造触发幻觉的查询。我们引入了 QueryBandits，这是一个赌博机框架，用于设计重写策略以最大化奖励模型，该模型基于输入查询的 17 个语言特征的敏感性来封装幻觉倾向，从而主动引导 LLMs 远离生成幻觉。在 13 个不同的问答基准和每个数据集 1,050 个经过词汇扰动的查询上，我们表现最好的上下文 QueryBandit（汤普森采样）相对于无重写基线实现了 87.5% 的胜率，并且分别比零样本静态提示（“改写”或“扩展”）高出 42.6% 和 60.3%。因此，我们通过以查询重写形式进行的干预，实证验证了 QueryBandits 在缓解幻觉方面的有效性。有趣的是，某些静态提示策略——它们构成了当前查询重写文献中的相当一部分——比无重写基线具有更高的累积遗憾，这表明静态重写可能会加剧幻觉。此外，我们发现收敛后的每臂回归特征权重向量证实了不存在对所有查询都最优的单一重写策略。在此背景下，通过 QueryBandits 利用语义特征进行引导式重写，可以通过前向传播机制诱导输出行为的显著变化，从而无需重新训练或基于梯度的适应。
英文摘要：
Advanced reasoning capabilities in Large Language Models (LLMs) have caused higher hallucination prevalence; yet most mitigation work focuses on after-the-fact filtering rather than shaping the queries that trigger them. We introduce QueryBandits, a bandit framework that designs rewrite strategies to maximize a reward model, that encapsulates hallucination propensity based upon the sensitivities of 17 linguistic features of the input query-and therefore, proactively steer LLMs away from generating hallucinations. Across 13 diverse QA benchmarks and 1,050 lexically perturbed queries per dataset, our top contextual QueryBandit (Thompson Sampling) achieves an 87.5% win rate over a no-rewrite baseline and also outperforms zero-shot static prompting ("paraphrase" or "expand") by 42.6% and 60.3% respectively. Therefore, we empirically substantiate the effectiveness of QueryBandits in mitigating hallucination via the intervention that takes the form of a query rewrite. Interestingly, certain static prompting strategies, which constitute a considerable number of current query rewriting literature, have a higher cumulative regret than the no-rewrite baseline, signifying that static rewrites can worsen hallucination. Moreover, we discover that the converged per-arm regression feature weight vectors substantiate that there is no single rewrite strategy optimal for all queries. In this context, guided rewriting via exploiting semantic features with QueryBandits can induce significant shifts in output behavior through forward-pass mechanisms, bypassing the need for retraining or gradient-based adaptation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
