arXiv：cs.LG（机器学习，全量分类）· Shireen Kudukkil Manchingal, Sofiia Nikolenko, Fabio Cuzzolin·· 15 小时前AI 评分47
Credal 大语言模型：不确定性下的语义承诺
Credal Large Language Models for Semantic Commitment under Uncertainty
AI 导读
研究者提出 Credal 大语言模型（CLLMs），用一组 LoRA 适配器集成构建 credal 集，以概率上下界替代单一 softmax 输出，并据此制定承诺规则：仅当答案下界概率超过所有备选答案的上界概率时才作答。
正文
Abstract:Large language models (LLMs) often produce fluent but incorrect answers with unwarranted confidence. A central limitation is that standard LLMs represent uncertainty through a single predictive distribution, conflating epistemic ignorance with genuine ambiguity. We introduce Credal Large Language Models (CLLMs): an ensemble of LoRA adapters induces a credal set whose lower and upper probabilities expose the spread of plausible predictive distributions rather than collapsing to a single softmax output. From this representation, we derive a single commitment rule: the model commits to an answer only when its lower probability exceeds the upper probability of every alternative, and otherwise returns the set of answers that no plausible predictor rules out. We apply this commitment rule at two depths: Credal Token Commitment (CTC) applies it to answer tokens from one ensemble forward pass, which decides constrained answers without any generation; for open-ended answers, credal decoding extends a partial answer only when no completed answer dominates it, so that the completions produced are those the plausible predictors license, and Credal Semantic Commitment (CSC) applies the rule to their meaning clusters. We evaluate CLLMs with Gemma-2-9B, Llama-3.1-8B and Qwen2.5-7B on OpenBookQA, CoQA, TriviaQA and ARC-Challenge. On multiple choice, CTC commits on 73-91% of questions at 89-98% accuracy, returns sets of 1.1-1.5 options containing the gold one on 89-98%, and its intervals contain the observed accuracy in 24 of 30 confidence bins without calibration; corrupted context lowers commitment from 87-92% to 65-71%, and on Gemma the credal bound detects corruption better than every baseline. On open-ended QA, CLLM outperforms semantic entropy and Laplace-LoRA at a fixed coverage by up to 19% and 9.5% absolute accuracy on CoQA and TriviaQA with context, for every backbone.
| Comments: | 45 pages, 10 figures, 19 tables |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML) |
| MSC classes: | 68T37 |
| Cite as: | arXiv:2608.23244 [cs.CL] |
| (or arXiv:2608.23244v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2608.23244 arXiv-issued DOI via DataCite |
Submission history
From: Fabio Cuzzolin [view email]
[v1]
Mon, 24 Aug 2026 13:30:55 UTC (285 KB)
[v2]
Thu, 1 Oct 2026 12:30:38 UTC (620 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org