Pith. sign in

REVIEW 9 cited by

SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2405.20974 v3 pith:3IQ3PGYE submitted 2024-05-31 cs.CL cs.AIcs.LG

Tianyang Xu , Shujin Wu , Shizhe Diao , Xiaoze Liu , Xingyao Wang , Yangyi Chen , Jing Gao This is my paper · ORCID
classification cs.CLcs.AIcs.LG
keywords confidencellmssayselfestimatesrationalesself-reflectiveaccurateapproaches
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Large language models (LLMs) often generate inaccurate or fabricated information and generally fail to indicate their confidence, which limits their broader applications. Previous work elicits confidence from LLMs by direct or self-consistency prompting, or constructing specific datasets for supervised finetuning. The prompting-based approaches have inferior performance, and the training-based approaches are limited to binary or inaccurate group-level confidence estimates. In this work, we present the advanced SaySelf, a training framework that teaches LLMs to express more accurate fine-grained confidence estimates. In addition, beyond the confidence scores, SaySelf initiates the process of directing LLMs to produce self-reflective rationales that clearly identify gaps in their parametric knowledge and explain their uncertainty. This is achieved by using an LLM to automatically summarize the uncertainties in specific knowledge via natural language. The summarization is based on the analysis of the inconsistency in multiple sampled reasoning chains, and the resulting data is utilized for supervised fine-tuning. Moreover, we utilize reinforcement learning with a meticulously crafted reward function to calibrate the confidence estimates, motivating LLMs to deliver accurate, high-confidence predictions and to penalize overconfidence in erroneous outputs. Experimental results in both in-distribution and out-of-distribution datasets demonstrate the effectiveness of SaySelf in reducing the confidence calibration error and maintaining the task performance. We show that the generated self-reflective rationales are reasonable and can further contribute to the calibration. The code is made public at https://github.com/xu1868/SaySelf.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. MMBoundary: Advancing MLLM Knowledge Boundary Awareness through Reasoning Step Confidence Calibration

    cs.CL 2025-05 conditional novelty 6.0 of 10

    MMBoundary trains multimodal LLMs to express per-step natural-language confidence and uses reinforcement learning with three rewards to calibrate those statements, cutting calibration error by 7.5% and boosting task a...

  2. Antidote: A Unified Framework for Mitigating LVLM Hallucinations in Counterfactual Presupposition and Object Perception

    cs.CV 2025-04 conditional novelty 6.0 of 10

    Antidote trains LVLMs to reject false presuppositions in questions and reduce object hallucination using synthetic images, self-corrected preferences, and a new CP-Bench benchmark.

  3. Object-Level Verbalized Confidence Calibration in Vision-Language Models via Semantic Perturbation

    cs.CV 2025-04 conditional novelty 6.0 of 10

    A new training framework, CSP, uses semantic perturbation of object regions plus preference optimization to make VLMs verbalize confidence that better matches answer correctness.

  4. When One LLM Drools, Multi-LLM Collaboration Rules

    cs.CL 2025-02 conditional novelty 6.0 of 10

    A position paper that introduces a four-level taxonomy of multi-LLM collaboration (API, text, logit, weight) and argues it is essential for reliability, pluralism, and democratization.

  5. Evaluating Prompt Engineering Techniques for Accuracy and Confidence Elicitation in Medical LLMs

    cs.CY 2025-05 conditional novelty 5.0 of 10

    Across 156 configurations on Persian medical board questions, Chain-of-Thought prompting raised accuracy while increasing overconfidence, and emotional prompting inflated confidence without accuracy gains.

  6. Observing Micromotives and Macrobehavior of Large Language Models

    physics.soc-ph 2024-12 reject novelty 5.0 of 10

    Using Schelling's model, the authors find that LLM-generated move/stay advice leads to a similarly segregated society across models and demographic categories, regardless of the model's measured bias.

  7. Position: Uncertainty Quantification Needs Reassessment for Large-language Model Agents

    cs.LG 2025-05 conditional novelty 4.0 of 10

    A position paper arguing that aleatoric/epistemic uncertainty splits fail for LLM agents and proposing underspecification, interaction, and output-based uncertainty research.

  8. A Survey on Uncertainty Quantification of Large Language Models: Taxonomy, Open Research Challenges, and Future Directions

    cs.CL 2024-12 conditional novelty 4.0 of 10

    A review that organizes LLM uncertainty quantification into token-level, self-verbalized, semantic-similarity, and mechanistic interpretability categories.

  9. Loki's Dance of Illusions: A Comprehensive Survey of Hallucination in Large Language Models

    cs.CL 2025-06 reject novelty 3.0 of 10

    A survey of LLM hallucination research that formalizes hallucination types and argues, via incompleteness and undecidability arguments, that hallucinations cannot be fully eliminated.

Pith tools