Cover Image

Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yang, Zhe, Zhang, Yichang, Wang, Yudong, Xu, Ziyao, Lin, Junyang, Sui, Zhifang
Format:	Preprint
Published:	2024
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

A Probabilistic Inference Scaling Theory for LLM Self-Correction
by: Yang, Zhe, et al.
Published: (2025)

Can Large Language Models Always Solve Easy Problems if They Can Solve Harder Ones?
by: Yang, Zhe, et al.
Published: (2024)

Towards Better RL Training Data Utilization via Second-Order Rollout
by: Yang, Zhe, et al.
Published: (2026)

Enhancing Reliability across Short and Long-Form QA via Reinforcement Learning
by: Wang, Yudong, et al.
Published: (2025)

Exploring Activation Patterns of Parameters in Language Models
by: Wang, Yudong, et al.
Published: (2024)

Reducing Hallucinations in Entity Abstract Summarization with Facts-Template Decomposition
by: Zhu, Fangwei, et al.
Published: (2024)

Not All Demonstration Examples are Equally Beneficial: Reweighting Demonstration Examples for In-Context Learning
by: Yang, Zhe, et al.
Published: (2023)

Evaluating and Aligning CodeLLMs on Human Preference
by: Yang, Jian, et al.
Published: (2024)

On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept
by: Liu, Guangliang, et al.
Published: (2024)

Confidence Matters: Revisiting Intrinsic Self-Correction Capabilities of Large Language Models
by: Li, Loka, et al.
Published: (2024)

Investigating the (De)Composition Capabilities of Large Language Models in Natural-to-Formal Language Conversion
by: Xu, Ziyao, et al.
Published: (2025)

Supervised Optimism Correction: Be Confident When LLMs Are Sure
by: Zhang, Junjie, et al.
Published: (2025)

CriticBench: Benchmarking LLMs for Critique-Correct Reasoning
by: Lin, Zicheng, et al.
Published: (2024)

HellaSwag-Pro: A Large-Scale Bilingual Benchmark for Evaluating the Robustness of LLMs in Commonsense Reasoning
by: Li, Xiaoyuan, et al.
Published: (2025)

Meeseeks: A Feedback-Driven, Iterative Self-Correction Benchmark evaluating LLMs' Instruction Following Capability
by: wang, Jiaming, et al.
Published: (2025)

CoLT: Reasoning with Chain of Latent Tool Calls
by: Zhu, Fangwei, et al.
Published: (2026)

Language Models can Self-Lengthen to Generate Long Texts
by: Quan, Shanghaoran, et al.
Published: (2024)

DeepInnovator: Triggering the Innovative Capabilities of LLMs
by: Fan, Tianyu, et al.
Published: (2026)

From Mathematical Reasoning to Code: Generalization of Process Reward Models in Test-Time Scaling
by: Chen, Zhengyu, et al.
Published: (2025)

CritiCal: Can Critique Help LLM Uncertainty or Confidence Calibration?
by: Zong, Qing, et al.
Published: (2025)

SG-FSM: A Self-Guiding Zero-Shot Prompting Paradigm for Multi-Hop Question Answering Based on Finite State Machine
by: Wang, Xiaochen, et al.
Published: (2024)

Self-Evolving Critique Abilities in Large Language Models
by: Tang, Zhengyang, et al.
Published: (2025)

RMTBench: Benchmarking LLMs Through Multi-Turn User-Centric Role-Playing
by: Xiang, Hao, et al.
Published: (2025)

Dancing with Critiques: Enhancing LLM Reasoning with Stepwise Natural Language Self-Critique
by: Li, Yansi, et al.
Published: (2025)

Beyond Output Critique: Self-Correction via Task Distillation
by: Rahmani, Hossein A., et al.
Published: (2026)

Generalization v.s. Memorization: Tracing Language Models' Capabilities Back to Pretraining Data
by: Wang, Xinyi, et al.
Published: (2024)

Sparse-BitNet: 1.58-bit LLMs are Naturally Friendly to Semi-Structured Sparsity
by: Zhang, Di, et al.
Published: (2026)

Fact-Level Confidence Calibration and Self-Correction
by: Yuan, Yige, et al.
Published: (2024)

Self-Boosting Large Language Models with Synthetic Preference Data
by: Dong, Qingxiu, et al.
Published: (2024)

Chain-of-Thought Tokens are Computer Program Variables
by: Zhu, Fangwei, et al.
Published: (2025)

CRITICTOOL: Evaluating Self-Critique Capabilities of Large Language Models in Tool-Calling Error Scenarios
by: Huang, Shiting, et al.
Published: (2025)

Language Models Encode the Value of Numbers Linearly
by: Zhu, Fangwei, et al.
Published: (2024)

FSM: A Finite State Machine Based Zero-Shot Prompting Paradigm for Multi-Hop Question Answering
by: Wang, Xiaochen, et al.
Published: (2024)

Stabilizing MoE Reinforcement Learning by Aligning Training and Inference Routers
by: Ma, Wenhan, et al.
Published: (2025)

Self-Correction Makes LLMs Better Parsers
by: Zhang, Ziyan, et al.
Published: (2025)

CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing
by: Gou, Zhibin, et al.
Published: (2023)

Confidence Improves Self-Consistency in LLMs
by: Taubenfeld, Amir, et al.
Published: (2025)

Understanding the Dark Side of LLMs' Intrinsic Self-Correction
by: Zhang, Qingjie, et al.
Published: (2024)

When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMs
by: Kamoi, Ryo, et al.
Published: (2024)

SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales
by: Xu, Tianyang, et al.
Published: (2024)