MASH: Modeling Abstention via Selective Help-Seeking
LLMs cannot reliably recognize their parametric knowledge boundaries and often hallucinate answers to outside-of-boundary questions. In this paper, we introduce MASH (Modeling Abstention via Selective Help-seeking), a training framework that readily extracts abstentions from LLMs. Our key idea is that any external help-seeking by an LLM, i.e. search tool use, can serve as a proxy for abstention if the external help (search) is appropriately penalized while also rewarding answer accuracy. MASH operationalizes this idea using reinforcement learning with a pay-per-search reward. We run experiments on three knowledge-intensive QA datasets. Our results show that MASH substantially improves upon the selective help-seeking performance of prior efficient search approaches; on multi-hop datasets, it improves answer accuracy by 7.6%. Furthermore, MASH demonstrates strong off-the-shelf abstention performance, showcasing behavior competitive with prior abstention methods that additionally require predetermining model knowledge boundaries to construct training data. Overall, we show MASH training effectively aligns search tool use with parametric knowledge, which can be successfully leveraged for making abstention decisions and efficient search tool use
Code (0)
등록된 구현이 없습니다.
Tasks
Reinforcement LearningSimilar Papers 제목 키워드 기반
Online Selective Classification with Limited Feedback
Motivated by applications to resource-limited and safety-critical domains, we study selective classification in the online learning model, wherein a predictor may abstain from classifying an instance. For example, this m…
ClassificationvalidKnowing When to Quit: A Principled Framework for Dynamic Abstention in LLM Reasoning
LLMs utilizing chain-of-thought reasoning often waste substantial compute by producing long, incorrect responses. Abstention can mitigate this by withholding outputs unlikely to be correct. While most abstention methods …
Mathematical ReasoningReinforcement LearningCMasher: Scientific colormaps for making accessible, informative and 'cmashing' plots
CMasher is a Python package that provides a curated collection of scientific colormaps, showcased in the online documentation (https://cmasher.readthedocs.io). The colormaps in CMasher are all designed to be perceptually…
Knowing When Not to Predict: Self Supervised Learning and Abstention for Safer DR Screening
Self-supervised learning (SSL) is now a standard way to pretrain medical image models, but performance is still mostly judged by downstream accuracy. For safety-critical screening tasks such as diabetic retinopathy gradi…
Diabetic Retinopathy GradingSelf-Supervised LearningMonster Mash: A Single-View Approach to Casual 3D Modeling and Animation
Monster Mash is a new sketch-based modeling and animation tool that allows you to quickly sketch a character, inflate it into 3D, and promptly animate it. You can perform all interactions in the sketching plane. No 3D ma…
Image Generation