LaMsS: When Large Language Models Meet Self-Skepticism

Yetao Wu, Yihong Wang, Teng Chen, Ningyuan Xi, Qingqing Gu, Hongyang Lei, Luo Ji

Keywords: skepticism, LLM, pretrain, finetune, hallucination, self-awareness, consciousness

TL;DR: We propose a novel approach called LaMsS, which leverages self-skeptical thinking on LLMs to mitigate hallucination.

Abstract: Hallucination is a major challenge for large language models (LLMs), preventing their further application in some fields. The skeptical thinking of humankind could be useful for LLMs to self-cognition, self-reflection and alleviate their hallucinations. Inspired by this consideration, we propose a novel approach called LaMsS, which combines the semantic understanding capability of LLMs with self-skepticism. By introducing a series of skepticism tokens and augmenting them into the vocabulary, we conduct both pertaining and finetuning, which allow the LLM to decode each normal token followed by a skeptical token, representing different skepticism levels. By calculating the response skepticism given a query, one can define a new self-aware LLM which is only willing to answer with relative lower skepticism level than the threshold. By examining the accuracy, AUC and AP of willingly answering questions, we demonstrate that LaMsS achieves better performance than baselines on both multi-choice questions and open-domain question-answering benchmarks, and can generalize to multi-task and out-of-domain settings. Our study sheds some lights on the self-skepticism modeling on further artificial intelligence. Project code and model checkpoints can be found in https://anonymous.4open.science/r/SM-1E76.

Submission Number: 65