When Models Hesitate: Answer Instability as a Label-Free Uncertainty Signal for LLMs (2026.acl-srw)
Copied to clipboard
Jasper Meynard Arana, Kristine Ann M. Carandang, Ethan Robert Casin, Christian Alis, Christopher Monterola
| Challenge: | Existing approaches to uncertainty estimation typically require access to internals, additional supervision, or computationally intensive pipelines. |
| Approach: | They propose to use a label-free uncertainty signal to predict the variability of a model's final answer across repeated stochastic generations of the same prompt to achieve performance competitive with semantic entropy. |
| Outcome: | The proposed method achieves performance competitive with semantic entropy while requiring no similarity model. |
Similar Papers
Adaptive Prompt Optimization for Open-Ended Tasks: Uncertainty Preference as a Secondary Signal (2026.findings-acl)
Copied to clipboard
| Challenge: | Recent training-free prompt optimizers treat performance as maximizing a single scalar score and ignore a second signal that the desired style is task dependent. |
| Approach: | They propose a semantic-entropy-based method that uses task uncertainty to guide prompt optimization by selecting high-entropicy candidates for creative tasks and low-energetic candidates for conservative ones. |
| Outcome: | The proposed method outperforms baselines on MT-Bench subsets and integrates easily into existing prompt optimizers. |
Rethinking LLM Uncertainty: A Multi-Agent Approach to Estimating Black-Box Model Uncertainty (2025.findings-emnlp)
Copied to clipboard
Yu Feng, Phu Mon Htut, Zheng Qi, Wei Xiao, Manuel Mager, Nikolaos Pappas, Kishaloy Halder, Yang Li, Yassine Benajiba, Dan Roth
| Challenge: | Existing methods to gauge model’s uncertainty through self-consistency in responses to the target query are misleading: an LLM may confidently provide an incorrect answer to a target query, yet give a confident and accurate answer to that same query when answering a knowledge-preserving perturbation of the query. |
| Approach: | They propose a method that uses multi-agent interaction to estimate black-box LLMs' uncertainty. |
| Outcome: | The proposed method outperforms existing self-consistency based methods and improves hallucination detection. |
Calibrating LLM Confidence by Probing Perturbed Representation Stability (2025.emnlp-main)
Copied to clipboard
Reza Khanmohammadi, Erfan Miahi, Mehrsa Mardikoraem, Simerjot Kaur, Ivan Brugere, Charese Smiley, Kundan S Thind, Mohammad M. Ghassemi
| Challenge: | Despite their impressive performance, large language models (LLMs) consistently struggle with confidence calibration. |
| Approach: | They propose a method to analyze internal representational stability in large language models by applying adversarial perturbations to final hidden states and using a lightweight classifier to predict answer correctness. |
| Outcome: | CCPS significantly outperforms existing methods on LLMs from 8B to 32B parameters in multiple-choice and open-ended formats. |
Prompting the Unknown: Understanding Response Uncertainty in Large Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Large language models are widely used in decision-making across diverse domains. |
| Approach: | They propose a prompt-response concept model that explains the relationship between the amount of task-relevant information provided in the prompt and the LLM-generated response uncertainty by identifying four sources of response uncertainty. |
| Outcome: | The proposed model shows that the amount of information provided in the prompt influences the LLM-generated response uncertainty. |
Bayesian Prompt Ensembles: Model Uncertainty Estimation for Black-Box Large Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to quantify uncertainty of pre-trained large language models require specific architectures or retraining strategies. |
| Approach: | They propose a Bayesian Prompts Ensemble approach to accurately quantify LLM uncertainty . they compute output probabilities through a weighted ensemble of different task instruction prompts . |
| Outcome: | The proposed approach achieves significantly superior calibration over baselines over a range of natural language classification tasks. |
“My Answer is C”: First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models (2024.findings-acl)
Copied to clipboard
Xinpeng Wang, Bolei Ma, Chengzhi Hu, Leon Weber-Genzel, Paul Röttger, Frauke Kreuter, Dirk Hovy, Barbara Plank
| Challenge: | Multiple choice questions are one of the most popular evaluation formats for understanding the capabilities of autoregressive large language models (LLMs). |
| Approach: | They evaluated how aligned first-token evaluation is with the text output along several dimensions, namely final option choice, refusal rate, choice distribution and robustness under prompt perturbation. |
| Outcome: | The proposed evaluation methods are misaligned on all dimensions, reaching mismatch rates over 60%. |
Can LLMs Learn Uncertainty on Their Own? Expressing Uncertainty Effectively in A Self-Training Manner (2024.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) exhibit excessive, random, and uninformative uncertainty rendering them unsuitable for decision-making in human-computer interactions. |
| Approach: | They propose an uncertainty-aware instruction tuning method that aligns LLMs’ perception with the probabilistic uncertainty of the generation. |
| Outcome: | The proposed method improves LLMs' performance by 45.2%, with reasonably good out-of-domain generalization capabilities. |
From Confidence to Collapse in LLM Factual Robustness (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing evaluation methods focus on performance-based metrics, often investigating from the perspective of prompt perturbations, which captures only the externally triggered side of knowledge robustness. |
| Approach: | They propose a method to measure factual robustness from the perspective of the generation process by analyzing token distribution entropy and temperature scaling sensitivity. |
| Outcome: | The proposed method measures factual robustness from the perspective of the generation process and entropy and temperature scaling sensitivity. |
Reliable Gradient-free and Likelihood-free Prompt Tuning (2023.findings-eacl)
Copied to clipboard
| Challenge: | Large pre-trained language models are often offered as black-box APIs due to privacy or commercial constraints. |
| Approach: | They propose to tune the soft prompts without requiring gradient computation and extend the model to include a distribution over prompts. |
| Outcome: | The proposed methods are competitive with gradient-based approaches with full access to the PLM. |
Detecting LLM Hallucination Through Layer-wise Information Deficiency: Analysis of Ambiguous Prompts and Unanswerable Questions (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) often generate confident yet inaccurate responses, introducing significant risks for deployment in safety-critical domains. |
| Approach: | They propose a method to detect model hallucination by systematic analysis of information flow across model layers. |
| Outcome: | The proposed approach improves model reliability by immediately integrating with universal LLMs without additional training or architectural modifications. |