# When Should a Language Model Trust Itself? Same-Model Self-Verification as a Conditional Confidence Signal

arXiv: https://arxiv.org/abs/2605.02915 · HTML: https://mlsanity.com/paper/2605.02915/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 12 references that are in the map.
- First posted on arXiv 8 Apr 2026.

## Builds on (most cited first, 10 of 10)

- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs](https://mlsanity.com/paper/2306.13063/) (2023)
- [Teaching Models to Express Their Uncertainty in Words](https://mlsanity.com/paper/2205.14334/) (2022)
- [Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs](https://mlsanity.com/paper/2406.15927/) (2024)
- [Why Language Models Hallucinate](https://mlsanity.com/paper/2509.04664/) (2025)
- [On Verbalized Confidence Scores for LLMs](https://mlsanity.com/paper/2412.14737/) (2024)
- [Self-Evaluation Improves Selective Generation in Large Language Models](https://mlsanity.com/paper/2312.09300/) (2023)
- [Uncertainty in Natural Language Processing: Sources, Quantification, and Applications](https://mlsanity.com/paper/2306.04459/) (2023)

## Built on by (most cited first, 1 of 1)

- [Learning from Research: Toward Lifelong Agent Harness Evolution](https://mlsanity.com/paper/2609.40169/) (2026)

## Abstract (arXiv)

Same-model self-verification, prompting a model to audit its own predicted answer, is a plausible confidence signal for selective prediction, but its practical value remains unclear once strong likelihood-based baselines are taken seriously. We evaluate self-verification against two such baselines, LL-AVG and LL-SUM, on ARC-Challenge and TruthfulQA-MC across multiple model families, scales, and prompt variants. We measure not only correctness ranking, but also abstention quality through AURC and operating-point analyses. The results are sharply task- and model-dependent. On ARC-Challenge, self-verification substantially improves over LL-AVG for Phi-2 and the Qwen models, with the largest gains appearing in Qwen-7B. On TruthfulQA-MC, however, the signal is less reliable: smaller models can become prompt-sensitive, DeepSeek-R1-Distill-8B degrades relative to LL-AVG, and LL-SUM often remains the stronger practical baseline. We therefore do not treat self-verification as a general-purpose uncertainty estimator. In this setting, it is better understood as a conditional confidence signal whose value depends on task type, model family, prompt formulation, and, crucially, the baseline it must beat.