Can a Model Catch Its Own Hallucinations for Free?: Label-Free Doubt Signals Hold Their Own Against a Labelled Dataset for Abstention
Source
Ali Asaria, Tony Salomone, Deep Gandhi
Author
Ali Asaria, Tony Salomone, Deep Gandhi
Date
Terms in this piece · Glossary
fine-tuning — Taking a trained model and training it a bit more on your own examples so it gets better at one specific job.
open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
Why it matters
Abstention tuning may not require a labelled right/wrong dataset — a model's own tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → confidence is a free supervision signal that held its own against labelled baselines at matched coverage on short-form factual QA.
articleSelf- and Other-Labels Induce Bidirectional Bias in LLM JudgesSongeun Chae, Min Kim, Donghoon Jung, Seojin Choi, Seohyon Jung
articleMeasure, Don't Optimize: Forecasting Recovery in LLM UnlearningZirui Song, Huaxing Liu, Xiang Wang, Shuai Li, Xinye Li, Lang Gao, Jinghui Zhang, Zheng Lu, Fengxian Ji, Xiaojun Chang, Xiuying Chen
articleDiagnostic Foundation for Evaluating LLMs' Research Integrity as Co-ScientistsYash Tripathi, Silu Sharma, Sai Sidhanth Manoharan Jayanthi, Shivank Garg, Lin Li