LLM Self-Correction

LLM: Large Language Model

This field asks whether language models can recognize faulty claims, reasoning steps, or actions and revise them, which matters for reliability in tasks such as mathematics, coding, question answering, and clinical workflows. Approaches include prompted reflection and critique, feedback from code execution or formal verifiers, retrieval and multi-agent review, training with reinforcement learning on failed attempts, and signals from model uncertainty or internal representations. Evaluations use settings such as GSM8K and AIME for mathematics, HumanEval and SWE-bench for code, and Text-to-SQL benchmarks, often tracking both errors fixed and correct answers damaged. A central tension is that detecting an error does not ensure a useful repair: noisy feedback, poor error localization, or repeated revision can worsen outputs, motivating selective, evidence-grounded correction and reliable stopping rules.

Latest papers

159 papers24 papers in the last four weeks, up 167% on the four weeks before. 0.1% of all new papers.

Beyond Refusal: Probing the Limits of Agentic Self-Correction for Semantic Sensitive Information
Feb 25, 2026Umid Suleymanov, Zaur Rajabov, Emil Mirzazada +1Privacy Leakage in Language Models LLM Self-Correction
Department of Computer Science, Virginia Tech · School of IT and Engineering, ADA University.
AgenticSum: An Agentic Inference-Time Framework for Faithful Clinical Text Summarization
Feb 23, 2026Fahmida Liza Piya, Rahmatollah BeheshtiText Summarization LLM Hallucination Mitigation
University of Delaware, Newark, DE, USA
Think $^{2}$ : Grounded Metacognitive Reasoning in Large Language Models
Feb 21, 2026Abraham Paul Elenjical, Vivek Hruday Kavuri, Vasudeva VarmaLanguage Model Self-Assessment Metacognition in Language Models
IIIT Hyderabad
When Actions Go Off-Task: Detecting and Correcting Misaligned Actions in Computer-Use Agents
Feb 9, 2026Yuting Ning, Jaylen Jones, Zhehao Zhang +5LLM Self-Correction Computer-Use Agents
The Ohio State University · Amazon AGI
Agentic AI for Commercial Insurance Underwriting with Adversarial Self-Critique
Jan 21, 2026Joyjit Roy, Samaresh Kumar SinghLLM Self-Correction AI Agent Safety
Austin, Texas, USA · Leander, Texas, USA
Stepwise Think-Critique: Interleaved Reasoning and Self-Critique in a Single LLM
Dec 17, 2025Jiaqi Xu, Cuiling Lan, Xuejin Chen +1LLM Self-Correction CoT Reasoning
University of Science and Technology of China · ⋆This work was done when Jiaqi was an intern at MSRA. · Microsoft Research Asia
Superficial Reflection or Genuine Thought? A Fine-Grained Cognitive Analysis of Large Reasoning Models
Nov 30, 2025Yuxiang Chen, Zuohan Wu, Ziwei Wang +6LLM Self-Correction Cognitive Modeling
University College London · AI Lab, the Yangtze River Delta, China · The Hong Kong University of Science and Technology (Guangzhou) +4
Knowing When to Critique: Task-Adaptive Metacognitive Regulation for Reliable LLM Reasoning
Jul 20, 2025Xinmeng Hou, Ziting Chang, Zhouquan Lu +5LLM Self-Correction Multi-Agent LLMs
Nanyang Technological University · Unicorn Verse · Shanghai Jiao Tong University +3
SG-FSM: A Self-Guiding Zero-Shot Prompting Paradigm for Multi-Hop Question Answering Based on Finite State Machine
Oct 22, 2024Xiaochen Wang, Liang Chen, Reza Haf Zhe Yang +5Multi-Hop QA LLM Self-Correction
School of Software & Microelectronics, Peking University · International Digital Economy Academy · National Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University +1

Feb 25, 2026J/K move · Enter open · S save

Beyond Refusal: Probing the Limits of Agentic Self-Correction for Semantic Sensitive Information

Umid Suleymanov, Zaur Rajabov, Emil Mirzazada, Murat Kantarcioglu

Department of Computer Science, Virginia Tech · School of IT and Engineering, ADA University.

Privacy Leakage in Language Models LLM Self-Correction Privacy-Preserving Language Models Privacy-Utility Tradeoff

Open paper AskPDF arXiv

While defenses for structured PII are mature, Large Language Models (LLMs) pose a new threat: Semantic Sensitive Information (SemSI), where models infer sensitive identity attributes, generate reputation-harmful content, or hallucinate potentially wrong information. The capacity of LLMs to self-regulate these complex, context-dependent sensitive information leaks without destroying utility remains an open scientific question. To address this, we introduce SemSIEdit, an inference-time framework where an agentic "Editor" iteratively critiques and rewrites sensitive spans to preserve narrative flow rather than simply refusing to answer. Our analysis reveals a Privacy-Utility Pareto Frontier, where this agentic rewriting reduces leakage by 34.6% across all three SemSI categories while incurring a marginal utility loss of 9.8%. We also uncover a Scale-Dependent Safety Divergence: large reasoning models (e.g., GPT-5) achieve safety through constructive expansion (adding nuance), whereas capacity-constrained models revert to destructive truncation (deleting text). Finally, we identify a Reasoning Paradox: while inference-time reasoning increases baseline risk by enabling the model to make deeper sensitive inferences, it simultaneously empowers the defense to execute safe rewrites.