LLM Self-Correction

LLM: Large Language Model

Follow this topic Ask about this topic

Momentum

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

Jul 13Week of Sep 28

Latest papers 105

Jun 4, 2026cs.AI

Agentic Molecular Recovery via Molecule-Aware Exploration

Text-guided molecular generation with LLMs often yields invalid SMILES. We argue that invalid drafts should be addressed through a shift from validity-oriented repair to identity-preserving molecular recovery: the objective is not only to restore chemical validity, but also to preserve target-relevant structural cues and recover the molecular identity implied by the description. This perspective reveals the limitations of existing correction strategies. Post-hoc repair can recover validity while distorting key structures, LLM-only correction can introduce unintended global drift, and generic agentic correction remains constrained by greedy single-candidate trajectories even when equipped with executable RDKit edit tools. To address these limitations, we propose AMREC, which couples molecule-aware mismatch tracking with expanded candidate exploration and trajectory-level selection. On invalid ChEBI-20 drafts from three backbone models, AMREC achieves the strongest overall recovery profile across structural, exact-match, and string-level metrics.

Suwan Yoon, Changhee Lee

Department of Artificial Intelligence, Korea University / Seoul, Korea

LLM Self-Correction Molecular Generation

Jun 4, 2026cs.AI

Critic-Guided Heterogeneous Multi-Agent Reasoning for Reliable Mathematical Problem Solving

Recent Large Language Models (LLMs) have shown impressive reasoning abilities; but they are still susceptible to hallucinations, intermediate reasoning mistakes, and unreliable reasoning results in complex mathematical reasoning problems. In this study, we introduce a critic-based heterogeneous multi-agent approach to improve the dependability of mathematical reasoning. This framework incorporates several LLM agents of different specialties and employs a critic-driven adaptive learning system to assess and guide the reasoning process based on intermediate feedback. The system adopts a generator-validator framework, with the validator not only determining correctness but also offering critiques to guide regeneration of solutions. This allows for adaptive error correction and prevents error cascading. Our experiments on the GSM8K benchmark show that the proposed method achieves up to 13% accuracy improvement over single-shot and non-critic models. Additionally, findings suggest that heterogeneity and critique reduce the need for large models, allowing smaller models to perform on par. Ablation studies reveal the main performance gains are due to the critic-based feedback loop and not model size. In summary, the proposed approach showcases the benefits of combining heterogeneous multi-agent collaboration and critique to obtain reliable and interpretable reasoning systems.

Muhammad Talha Sharif, Abdul Rehman

dept. of Computer Science · National University of Computer and Emerging Sciences · Islamabad, Pakistan

LLM Self-Correction Multi-Agent Reasoning LLM Mathematical Reasoning LLM Agent Reliability

Jun 3, 2026cs.SE

Self-Reflective APIs: Structure Beats Verbosity for AI Agent Recovery

When an AI agent calls an API and hits a validation error, it needs more than what went wrong -- it needs what to do next. A self-reflective API returns, on validation failure, a machine-readable recovery_feedback.suggestions[] payload sufficient for the agent to repair the request and retry without external reasoning. On a leak-audited pilot (

N{=}30

per cell, 3 LLMs, 10 adversarial tasks), structured suggestions lift task-completion rate by

+36.7

--

40.0

pp over plain-English diagnoses on Anthropic models (Fisher's exact

p \le 0.0022

), at

1.8

--

2.2\times

better per-success token efficiency. The lift is not significant on gpt-4o-mini (

p{=}0.435

); a second-domain replication on a billing API confirms the pattern. The comparison only holds after auditing two undocumented classes of answer leakage in LLM benchmarks. We shipaudit_prompt_leakage.py as reusable CI infrastructure. Code and data: https://github.com/arquicanedo/self-reflective-apis.

Arquimedes Canedo, Grama Chethan

Siemens Digital Industries Software, USA

LLM Self-Correction Tool-Use Evaluation

Jun 3, 2026cs.CL

Imbuing Large Language Models with Bidirectional Logic for Robust Chain Repair

Autoregressive chain-of-thought (CoT) reasoning in large language models (LLMs) is fundamentally forward-directed: each step conditions only on prior tokens. This unidirectional inductive bias renders even capable models susceptible to error snowballing, wherein a single logical or arithmetic mistake in an early step irreversibly corrupts the entire reasoning chain. We introduce Teleological Reasoning Infilling (\TRI{}), a training framework that endows decoder-only transformers with a native \emph{goal-conditioned bridging} capability. The key insight is to reframe erroneous reasoning segments as fill-in-the-middle (FIM) tasks: given a verified prefix premise

P

, a verified downstream milestone

S

, and the original query

Q

, the model must synthesise the logical bridge

M

that connects

P

to

S

rigorously and completely. To achieve this with standard causal architectures, we introduce a Prefix-Suffix-Middle (PSM) sequence rearrangement with three non-overlapping sentinel tokens, enabling

M

to attend to both

P

and

S

without any structural modification to the self-attention mechanism. Training proceeds in two stages: (i) Supervised Fine-Tuning (SFT) on symbolically verified

(P, S, M)

triples extracted from formal mathematics corpora, and (ii) Direct Preference Optimisation (DPO) with a deterministic symbolic verifier (Lean 4 / Python) as the sole reward oracle, eliminating LLM-judge sycophancy. At inference, TRI operates as a surgical repair module within a dual-system loop: a causal draft model generates an initial trace, the verifier pinpoints failures, and TRI infills only the damaged segment, leaving verified sections intact. Comprehensive experiments on three benchmarks demonstrate that TRI achieves state-of-the-art performance across all tasks, while reducing per-problem token expenditure by 31.2%.

Zehua Cheng, Wei Dai, Jiahao Sun +1

Department of Computer Science, University of Oxford, UK · Institute of Logic and Computation, TU Wien, Austria

LLM Self-Correction Symbolic Reasoning Mathematical Reasoning Structured Reasoning LLM Reliability Efficient Language Model Inference+1

Jun 2, 2026cs.CR

Learn from Your Mistakes: Tree-like Self-Play for Secure Code LLMs

While Large Language Models (LLMs) excel in code generation, they remain prone to replicating subtle yet critical vulnerabilities endemic to their training data. Current alignment techniques, such as Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), typically apply coarse-grained optimization at the sequence level. This approach often fails to address the localized nature of security flaws, where a single incorrect token choice can compromise an entire program. To bridge this gap, we introduce Tree-like Self-Play (TSP), a framework that reframes secure code generation as a fine-grained sequential decision process. Unlike standard methods that blindly maximize likelihood, TSP constructs a decision tree where the model explores branching trajectories--generating both secure "golden paths" and vulnerable variants. By treating code generation as a self-play game, the model learns to strictly discriminate against its own localized errors. This provides a dense, on-policy learning signal that forces self-correction precisely at the critical decision nodes where vulnerabilities typically emerge. Our experiments demonstrate that TSP fundamentally enhances model reliability. In Python security benchmarks, TSP boosts CodeLlama-7B's pass rate (SPR@1) to 75.8%, significantly outperforming SFT (57.0%) and unstructured self-play baselines. Crucially, TSP induces robust out-of-distribution generalization: the model not only reduces vulnerabilities in unseen categories (CWEs) by 24.5% but also successfully transfers security principles learned from C/C++ to diverse languages, including Python, Go, and JavaScript. This suggests that TSP does not merely memorize patches, but internalizes abstract, language-agnostic security logic.

Wenqi Chen, Ziyan Zhang, Bin Wang +3

University of Electronic Science and Technology of China, Chengdu, China · Peking University, Beijing, China · University of Science and Technology of China, Hefei, China +2

RL for Code Generation LLM Self-Correction Code Generation Secure Code Generation

May 31, 2026cs.CL

Med-HEAL: Analyzing and Mitigating Hallucinations in Medical LLMs with Hallucination-Aware In-Context Learning

Hallucinations in medical large language models (LLMs) pose serious risks for clinical decision support, particularly when models must reason over complex electronic health records (EHRs). However, existing benchmarks often lack a realistic clinical context and provide limited insight into how hallucinations can be mitigated in practice. We introduce Med-HEAL, a framework for systematically identifying, analyzing, and mitigating hallucinations in medical LLMs using clinically grounded data. Building on the EHRNoteQA benchmark derived from MIMIC-IV discharge summaries, we construct a hallucination dataset by evaluating BioMistral-7B on open-ended clinical question answering tasks. Model outputs are labeled through a dual evaluation pipeline that combines LLM-as-a-Judge assessment (GPT-4o) with human auditing by medical student reviewers, producing correctness judgments and annotations of reasoning errors via a custom web-based evaluation system. We then leverage this dataset to investigate mitigation strategies: a self-critique pipeline, in which the test model reviews its own answers to detect potential errors and regenerates responses for flagged cases, and retrieval-augmented in-context learning (RA-ICL), which exposes the model to hallucinated and corrected examples. Experiments across five open-source LLMs-BioMistral, Llama-3.1, DeepSeek, Qwen2.5, and Qwen3, show that the self-critique strategy improves accuracy for three of five models (p < 0.05) without requiring parameter updates. Med-HEAL provides both a reusable hallucination dataset and a practical framework for studying and mitigating hallucinations in medical LLMs, supporting safer deployment of AI systems in clinical environments. Our code and data are publicly available at https://github.com/yimingliao-blad/med-heal.git.

Yiming Liao, Zeno Franco, Jose Eduardo Lizarraga Mazaba +1

University of Maryland, Baltimore County Baltimore, Maryland, USA · Concordia University Wisconsin · Medical College of Wisconsin Milwaukee, Wisconsin USA

LLM Self-Correction LLM Hallucination Mitigation Clinical QA In-Context Learning LLM Hallucination Detection Retrieval-Augmented ICL+1

May 28, 2026cs.SE

REPOT: Recoverable Program-of-Thought via Checkpoint Repair

One-shot Program-of-Thought (PoT) emits a Python program that prints a primitive-action plan; a single invalid action silently invalidates the trajectory. We introduce RePoT (Recoverable PoT): a deterministic verified replay that walks the plan through the environment to its first invalid transition, then one LLM call that resumes from the verified prefix. RePoT costs at most one extra LLM call on the ~14% of problems where PoT fails. RePoT beats PoT by +3 to +11pp across four closed-model configurations on PuzzleZoo-775 and peaks at 96.9% vs 86.3% on gpt-5.4-mini-medium; against the matched-budget PoT-retry baseline, RePoT wins decisively on Gemini (+3.8pp, 95% CI [+2.2,+5.4]), is within sampling noise on GPT-medium and Claude, and loses on GPT-mini -- a capability-scaling pattern we begin to address with Adaptive RePoT, a rule-based dispatcher that routes between suffix repair and a fresh PoT retry based on verified-prefix length (preliminary). We replicate on PlanBench Blocksworld (+1.1 to +11.4pp) and on four open-weights models (+3.3 to +20.0pp on three of four). On Derail-550, our controlled recovery benchmark, every condition with access to checkpoint information clears >=30% on GPT-medium and >=70% on Gemini, vs <=3.1% for error-only feedback -- showing that checkpoint information, not the specific verified-prefix tail, is the load-bearing recovery signal.

Parsa Mazaheri

University of California, Santa Cruz

LLM Self-Correction LLM-Based Program Synthesis Model-Based Planning

May 28, 2026cs.CL

CRITIC-R1: Learning Structured Critics for Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) improves knowledge-intensive question answering by incorporating external evidence. However, existing RAG methods still suffer from hallucinations and subtle reasoning errors. Recent studies introduce external critics to refine RAG outputs, yet they often provide coarse-grained and weakly structured feedback, exhibit over-aggressive intervention, and lead to noisy and unreliable refinement, limiting their effectiveness for correction. To tackle these issues, we propose CRITIC-R1, a structured critic framework that formulates and learns RAG critique as an explicit error diagnosis problem using reinforcement learning (RL). Our framework categorizes common RAG errors into multiple diagnostic dimensions, including verdict, error location, reasoning analysis, and fix generation. To learn these capabilities, we design two reward functions: Conservative Judgement Alignment (CJA) first encourages calibrated high-level judgements while mitigating the over-aggressive phenomenon, whereas Diagnostic Quality Alignment (DQA) further improves fine-grained diagnostic feedback through gated rewards. We train the critic model using GRPO-based RL with process-level supervision collected from external LLM teacher models. Experiments across five QA benchmarks show that CRITIC-R1 consistently improves answer quality over strong RAG baselines. Our source code is available at https://anonymous.4open.science/r/critic-r1-FCB0

Wenhan Xiao, Ziwei Zhang, Chuanyue Yu +4

Nankai University · Beihang University · Guangxi Normal University

LLM Self-Correction Retrieval-Augmented Generation

May 27, 2026cs.AI

DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes

Reinforcement learning has become a central paradigm for advancing reasoning in large language models, yet most existing methods still depend on stronger teacher models or heavily curated difficult datasets, limiting scalable capability improvement. In this paper, we introduce DenoiseRL, a reinforcement learning framework that substitutes external supervision with recovery-oriented optimization over failures from weak models. Instead of relying on stronger supervision or carefully engineered data, DenoiseRL learns directly from incorrect reasoning traces by converting them into opportunities for improvement, making training more scalable and less dependent on external resources. This yields a richer and more diverse learning signal, improving exploration efficiency from imperfect model behavior. As a result, DenoiseRL improves reasoning performance and overall training efficiency while reducing the need for expensive data curation or stronger teacher models. Empirically, DenoiseRL consistently outperforms strong on-policy RL baselines across competitive mathematical and general reasoning benchmarks and promotes stronger self-corrective behavior as training difficulty increases, highlighting an effective and scalable alternative pathway for improving reasoning in large language models.

Caijun Xu, Changyi Xiao, Zhongyuan Peng +1

1Fudan University · 2Shanghai Innovation Institute

LLM Self-Correction RL for Language Model Reasoning RL Post-Training RL Fine-Tuning

May 26, 2026cs.MA

Detection Without Correction: A Two-Parameter Decomposition of Multi-Stage LLM Pipelines

Multi-stage LLM pipelines that perform multi-agent debate, intrinsic self-correction, or retrieval-augmented verification exhibit puzzling aggregate behaviors: accuracy plateaus and reversals across rounds, non-replication of debate gains on contemporary frontier models, intrinsic self-correction degradation, and qualitative cross-provider divergence in debate dynamics. Downstream agent response can be operationalized as two coupled decisions: detection (whether to treat upstream content as authoritative) and conditional generation (what to produce if not). This decomposition yields four observable response regimes, of which detection-without-correction is the load-bearing failure mode. Across a nine-cell empirical grid spanning four model families, four benchmarks (GSM8K, MATH-500, GPQA-Diamond, AIME), and two methods (multi-agent debate, intrinsic self-correction), we find that the conditional miscorrection rate is consistently dominant (53-94% across cohorts) while detection rate varies contextually by more than an order of magnitude. The framework unifies the four phenomena above as signatures of a common mechanism and characterizes detection threshold as a stable model/protocol-level regularity that persists across methods at matched benchmark difficulty.

Prashanti Nilayam, Kiran Ramanna, Prashil Tumbade

Servicenow CA, USA

LLM Self-Correction LLM Evaluation Language Model Error Detection Multi-Agent Debate LLM Agent Reliability

May 23, 2026cs.CL

Guarded Repair for Harm-Aware Post-hoc Replacement of LLM Mathematical Reasoning

Post-hoc repair of LLM mathematical reasoning introduces an asymmetric risk: fixing an incorrect reasoning trace is useful, but replacing a trace that was already correct can be harmful. We study this problem under a selective replacement setting, where a system must decide whether a repaired candidate is safer than preserving the original cached trace. We present GuardedRepair, a guarded best-of-N repair framework that diagnoses cached reasoning traces, selectively triggers repair, and accepts answer-changing candidates only when deterministic verification guards support replacement. The framework combines lightweight symbolic checks, surface semantic-risk diagnostics, bounded candidate generation, and conservative acceptance policies. On the full GSM8K test set, where the initial reasoner already achieves 95.60% accuracy, GuardedRepair improves final accuracy to 96.89%, fixing 17 of 58 remaining errors without measured broken-correct cases in the main run. On a weak-reasoner ASDiv setting, accuracy improves from 78.40% to 87.60%. Direct regeneration baselines show that this gain is not explained by stronger-model re-solving alone: re-solving all GSM8K examples lowers accuracy to 93.03% and breaks 47 initially correct answers. Additional analyses show that guarded repair substantially improves the fixed/broken tradeoff, while also revealing that replacement risk is reduced rather than eliminated. These results support viewing post-hoc repair as harm-aware selective replacement rather than unconstrained re-solving.

Haizhou Xia

LLM Self-Correction LLM Mathematical Reasoning

May 20, 2026cs.MA

ProCrit: Self-Elicited Multi-Perspective Reasoning with Critic-Guided Revision for Multimodal Sarcasm Detection

Multimodal sarcasm detection requires reasoning over cross-modal incongruities between literal expression and intended meaning, yet the specific analytical perspectives needed vary across samples due to the diversity of sarcastic mechanisms. While recent methods make this analytical process explicit, they still rely on fixed, predefined perspectives that operate independently under hand-crafted routing rules. We argue that multimodal sarcasm detection instead calls for self-elicited multi-perspective reasoning, where a model autonomously generates the perspectives needed for each sample and progressively integrates them into a coherent analysis. To realize this goal, we propose ProCrit, a Proposal-Critic two-agent framework with a proposal agent for multi-perspective reasoning and a critic agent for external evaluation and targeted revision guidance. First, to overcome the lack of process-level supervision in existing sarcasm datasets, ProCrit synthesizes process-level reasoning annotations through a dynamic-role agentic rollout: a strong vision-language model sequentially spawns analytical roles within a shared context, and the resulting multi-role trajectories are flattened into sequences that preserve cross-perspective dependencies while enabling efficient autoregressive generation. Second, to improve reasoning reliability, ProCrit adopts a draft-critique-revise paradigm in which an independent critic identifies reasoning deficiencies and provides targeted natural-language feedback for directed revision. Finally, we develop a mutual-refinement training framework that jointly optimizes proposal drafting and feedback-guided revision via dual-stage reinforcement learning, while refining the critic agent according to the actual effectiveness of its feedback. Experiments on three widely used benchmarks demonstrate the effectiveness of ProCrit.

Yingjia Xu, Jiulong Wu, Bowen Zhang +3

School of Computer Science and Technology, Soochow University, Suzhou, China · Baidu Inc., Beijing, China · Zhipu AI, Beijing, China

LLM Self-Correction Multimodal CoT Reasoning

May 17, 2026cs.AI

CyberCorrect: A Cybernetic Framework for Closed-Loop Self-Correction in Large Language Models

Large language model (LLM) self-correction -- the ability to detect and fix errors in generated outputs -- remains largely ad hoc, relying on generic prompts such as "please reconsider your answer" without systematic error analysis or convergence guarantees. We propose CyberCorrect, a framework that formalizes LLM self-correction as a closed-loop control system grounded in cybernetic theory. The framework models the LLM generator as the plant and introduces a tri-modal Error Detector (combining self-consistency, verbalized confidence, and logic-chain verification) as the sensor. A type-directed Correction Controller generates targeted repair instructions based on diagnosed error categories, while a Convergence Judge determines iteration termination using stability criteria adapted from control theory. We further introduce three control-theoretic evaluation metrics -- convergence rate, overshoot rate, and oscillation rate -- that capture correction dynamics beyond final accuracy. Experiments on our constructed CyberCorrect-Bench (440 reasoning tasks with annotated error types and correction paths) show that CyberCorrect achieves 79.8% final accuracy, improving upon the best existing self-correction method by 6.2 percentage points, while reducing overshoot (erroneous over-correction) by 41% through its convergence control mechanism.

Yuning Wu, Yingmin Liu, Yang Shu

School of Software, Henan University, Kaifeng, China · 2Zhejiang University, Hangzhou, China

LLM Self-Correction LLM Evaluation Language Model Error Detection Language Model-Based Control

May 14, 2026cs.CL

Learning from Failures: Correction-Oriented Policy Optimization with Verifiable Rewards

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective paradigm for improving the reasoning capabilities of large language models. However, RLVR training is often hindered by sparse binary rewards and weak credit assignment, resulting in ambiguous optimization signals and underutilization of the useful information embedded in failed trajectories. To address this challenge, we propose Correction-Oriented Policy Optimization (CIPO), a simple and effective extension to RLVR that converts on-policy failed trajectories into correction-oriented supervision, without relying on any external signals. By jointly optimizing correction samples derived from the model's own failed attempts together with the standard RLVR objective, CIPO improves learning effectiveness while explicitly enhancing the model's ability to correct its own errors. Extensive experiments across 11 benchmarks spanning mathematical reasoning and code generation demonstrate that CIPO consistently and significantly outperforms strong baselines in both reasoning and correction performance. Moreover, CIPO yields stronger pass@K gains, indicating that it improves the model's intrinsic reasoning capacity rather than merely redistributing probability mass over existing correct answers.

Mengjie Ren, Jie Lou, Boxi Cao +6

Chinese Information Processing Laboratory, Institute of Software, Chinese Academy of Sciences · University of Chinese Academy of Sciences · Xiaohongshu Inc.

RL for Code Generation LLM Self-Correction RL for Language Model Reasoning Reinforcement Learning with Verifiable Rewards

May 10, 2026cs.LG

Your Simulation Runs but Solves the Wrong Physics: PDE-Grounded Intent Verification for LLM-Generated Multiphysics Simulation Code

Execution-based evaluation of LLM-generated code implicitly treats successful execution as a proxy for correctness. In scientific simulation, this proxy is insufficient: a generated input file can run, mesh, and converge while encoding governing equations that differ from the user's intent. We call this mismatch between intended physics and generated code the comprehension-generation gap. We instantiate this in MOOSE, where Kernel and BC objects map compositionally to weak-form residual terms, enabling deterministic reconstruction of the encoded PDE and comparison against an intended contract. We formalize this comparison as the Intent Fidelity Score (IFS), a structural metric covering governing terms, BCs, ICs, coefficients, and time scheme. Building on IFS, we develop a PDE-grounded refinement loop that uses deterministic violation reports to correct generated code iteratively. We evaluate on MooseBench, a 220-case multiphysics benchmark with PDE-level ground truth released with this work. On this benchmark, our method consistently improves mean IFS over direct generation, with gains concentrated on hard cases. On the subset where direct generation falls below IFS 0.7, refinement adds +0.22 to +0.41 absolute IFS. In the deployment audit, execution-only repair improves execution success while leaving 39-40% of all 220 cases runnable but still solving the wrong physics across the three main deployment-audit models, exposing executability and intent fidelity as separable failure modes. Static proof-of-concept experiments on four PDE-oriented DSLs (UFL/FEniCS, FreeFEM, FiPy, and Devito) suggest that the reconstruction-and-comparison pattern extends beyond MOOSE. These findings reinforce that executable simulation code should be verified against the mathematical structure it is intended to encode, not accepted on execution alone.

Zhenghan Song, Yulong Liu, Cheng Wan +4

Cornell University · Columbia University · Harvard University & Nanyang Technological University

LLM Self-Correction Scientific Code Generation Code Generation Evaluation Physics-Based Simulation

May 9, 2026cs.AI

Self-ReSET: Learning to Self-Recover from Unsafe Reasoning Trajectories

Large Reasoning Models possess remarkable capabilities for self-correction in general domain; however, they frequently struggle to recover from unsafe reasoning trajectories under adversarial attacks. Existing alignment methods attempt to mitigate this vulnerability by fine-tuning the model on expert data including reflection traces or adversarial prefixes. Crucially, these approaches are often hindered by static training data which inevitably deviate from model's dynamic, on-policy reasoning traces, resulting in model hardly covering its vast generation space and learning to recover from its own failures. To bridge this gap, we propose Self-ReSET, a pure reinforcement learning framework designed to equip LRMs with the intrinsic capacity to recover from their own safety error trajectories, which are subsequently reused as an initial state for reinforcement learning. Extensive experiments across various LRMs and benchmarks demonstrate that Self-ReSET significantly enhances robustness against adversarial attacks especially out-of-distribution (OOD) jailbreak prompts while maintaining general utility, along with efficient data utilization. Further analysis reveals that our method effectively fosters self-recovery patterns, enabling models to better identify and recover from unsafe intermediate error states back to benign paths. Our codes and data are available at https://github.com/Ing1024/Self-ReSET.

Dongcheng Zhang, Yi Zhang, Yuxin Chen +3

University of Science and Technology of China · Shanghai Artificial Intelligence Laboratory · National University of Singapore

LLM Self-Correction RL for Language Model Reasoning LLM Safety Adversarial Robustness of Language Models

May 7, 2026cs.LG

Knowing but Not Correcting: Routine Task Requests Suppress Factual Correction in LLMs

LLMs reliably correct false claims when presented in isolation, yet when the same claims are embedded in task-oriented requests, they often comply rather than correct. We term this failure mode \emph{correction suppression} and construct a benchmark of 300 false premises to systematically evaluate it across eight models. Suppression rates range from 19% to 90%, with four models exceeding 80%, establishing correction suppression as a prevalent and severe phenomenon. Mechanistic analysis reveals that suppression is not a knowledge failure: the model registers the error internally but task context diverts early-layer attention from the false claim as output intent crystallizes toward compliance at middle layers. We characterize this as \emph{knowing but not correcting} -- suppression occurs at response selection rather than knowledge encoding. Guided by this mechanism, we propose two training-free interventions. Correction Direction Steering (CDS) estimates a correction-compliance direction from matched pairs and injects it at middle layers before output intent crystallizes. Dynamic Payload Amplification (DPA) localizes payload tokens via attention divergence between early and late layers and amplifies their representation at the final layer, requiring no calibration data. Experiments on Qwen3.5-9B and LLaMA3.1-8B show both methods substantially improve factual strictness. CDS achieves the highest correction rate on Qwen3.5-9B (0%

\to

58.2%). DPA is the only method that preserves or improves reasoning capability on both models. These findings introduce \emph{factual strictness} -- the willingness to uphold accuracy against contextual pressures -- as a new dimension of model reliability.

Zixuan Chen, Hao Lin, Zizhe Chen +6

1Shanghai Jiao Tong University · 2The Chinese University of Hong Kong · 3Ant Group

LLM Self-Correction LLM Reliability Activation Steering

May 7, 2026cs.AI

ReFlect: An Effective Harness System for Complex Long-Horizon LLM Reasoning

Current reasoning paradigms for LLMs include chain-of-thought, ReAct, and post-hoc self-critique. These paradigms rely on two assumptions that fail on long-horizon, multi-stage tasks. As a result, errors accumulate silently across reasoning steps, leaving an open question: can a reasoning system effectively detect and recover from its own failures? We present ReFlect, a \emph{harness} system for LLM reasoning that creates standalone error detection and recovery logic as a deterministic wrapper around the model. Controlled experiments across 6 reasoning domains show that prompt-level self-critique produces formulaic templates that flag no issues in 90 of 100 audited reflection blocks, and the investigated LLMs wrongly accept a wrong answer in at least 76% of cases. Our ReFlect harness achieves task success rates ranging from 41% on gpt-4o-mini to 56% on Claude Sonnet 4.5 across six models spanning small and frontier scale, with per-model gains over Direct CoT ranging from +7 pp on Qwen2.5-72B to +29 pp on Claude Sonnet 4.5, and additionally raises SWE-bench patch-structural quality from 0% (Direct CoT) to between 82% (Qwen2.5-72B) and 87% (GPT-4o). Notably, the harness gain is inversely proportional to the model's Direct CoT task success rate (the fitted slope is -1.69 with r=-0.76): each pp lost in baseline success rate is mechanically recovered by 1.69 pp of harness gain. We spot that adding structured reasoning state and operators yields only 15.0--18.7% pair-mean on Llama-3.3-70B and Qwen2.5-72B because models at this scale cannot reliably populate the state its operators require. ReFlect is model-agnostic, training-free, and operates entirely at inference time.

Fan Huang

Indiana University Bloomington

LLM Self-Correction LLM Reliability LLM Reasoning

May 5, 2026cs.CL

The Counterexample Game: Iterated Conceptual Analysis and Repair in Language Models

Conceptual analysis -- proposing definitions and refining them through counterexamples -- is central to philosophical methodology. We study whether language models can perform this task through iterated analysis and repair chains: one model instance generates counterexamples to a proposed definition, another repairs the definition, and the process repeats. Across 20 concepts and thousands of counterexample-repair cycles, we find that, although many LM-generated counterexamples are judged invalid by both expert humans and an LM judge, the LM judge accepts roughly twice as many as humans do. Nonetheless, per-item validity judgments are moderately consistent across humans and between humans and the LM. We further find that extended iteration produces increasingly verbose definitions without improving accuracy. We also see that some concepts resist stable definitions in general. These findings suggest that while LMs can engage in philosophical reasoning, the counterexample-repair loop hits diminishing returns quickly and could be a fruitful test case for evaluating whether LMs can sustain high-level iterated philosophical reasoning.

Daniel Drucker, Kyle Mahowald

Department of Philosophy The University of Texas at Austin · Department of Linguistics The University of Texas at Austin

LLM Self-Correction LLM Evaluation Counterexample-Guided Refinement LLM Reasoning

May 5, 2026cs.AI

Revisiting the Travel Planning Capabilities of Large Language Models

Travel planning serves as a critical task for long-horizon reasoning, exposing significant deficits in LLMs. However, existing benchmarks and evaluations primarily assess final plans in an end-to-end manner, which lacks interpretability and makes it difficult to analyze the root causes of failures. To bridge this gap, we decompose travel planning into five constituent atomic sub-capabilities, including \emph{Constraint Extraction}, \emph{Tool Use}, \emph{Plan Generation}, \emph{Error Identification}, and \emph{Error Correction}. We implement a decoupled evaluation protocol leveraging oracle intermediate contexts to rigorously isolate these components, thereby measuring the atomic performance boundary without the noise of cascading errors. Our results highlight a clear contrast in performance: while LLMs are proficient in extracting explicit constraints, they struggle to infer implicit, open-world requirements. Furthermore, they exhibit structural biases in plan generation and suffer from ineffective self-correction, characterized by excessive sensitivity and erroneous persistence. These findings offer precise directions for improving LLM reasoning and planning abilities.

Bo-Wen Zhang, Jin Ye, Peng-Yu Hua +4

State Key Laboratory of Novel Software Technology, Nanjing University, China · School of Intelligence Science and Technology, Nanjing University, China · School of Artificial Intelligence, Nanjing University, China

LLM Self-Correction LLM Evaluation LLM Planning

May 1, 2026cs.SE

GeoContra: From Fluent GIS Code to Verifiable Spatial Analysis with Geography-Grounded Repair

Reliable spatial analysis in GIScience requires preserving coordinate semantics, topology, units, and geographic plausibility. Current LLM-based GIS systems generate fluent scripts but rarely enforce these geographic rules at scale. We present GeoContra, a verification and repair framework for LLM-driven Python GIS workflows. It represents each task as an executable geospatial contract-including natural-language questions, schemas, CRS metadata, expected outputs, spatial predicates, topology, metrics, required operations, and forbidden shortcuts. Generated programs undergo static rule inspection, runtime validation, and semantic verification, with violations fed back into a bounded repair loop. Evaluated on 7,079 real geospatial tasks across 15 Boston-area zones, 9 task families, and 11 open-source models (600 runs each), GeoContra improves spatial correctness on closed models from 47.6% to 77.5% for DeepSeek-V4 and from 57.7% to 81.5% for Kimi-K2.5. Across 11 open models, average correctness rises by 26.6%. GeoContra turns fluent code production into verifiable spatial analysis, catching negative travel times, CRS/field-schema violations, missing predicates, and brittle output casts that otherwise yield executable but geographically invalid results.

Yinhao Xiao, Rongbo Xiao, Yihan Zhang

School of Big Data and Artificial Intelligence, Guangdong University of Finance and Economics · School of Geography and Environment Economics, Guangdong University of Finance & Economics · Guangdong Engineering Research Center of Low-Altitude Remote Sensing Intelligent Monitoring +1

LLM Self-Correction Automated Program Repair Geospatial Reasoning Spatial Reasoning

Apr 30, 2026cs.AI

LLMs as ASP Programmers: Self-Correction Enables Task-Agnostic Nonmonotonic Reasoning

Recent large language models (LLMs) have achieved impressive reasoning milestones but continue to struggle with high computational costs, logical inconsistencies, and sharp performance degradation on high-complexity problems. While neuro-symbolic methods attempt to mitigate these issues by coupling LLMs with symbolic reasoners, existing approaches typically rely on monotonic logics (e.g., SMT) that cannot represent defeasible reasoning -- essential components of human cognition. We present "LLM+ASP," a framework that translates natural language into Answer Set Programming (ASP), a nonmonotonic formalism based on stable model semantics. Unlike prior "LLM+ASP" approaches that require manually authored knowledge modules, domain-specific prompts, or evaluation restricted to single problem classes, our framework operates without any per-task engineering and applies uniformly across diverse reasoning tasks. Our system utilizes an automated self-correction loop where structured feedback from the ASP solver enables iterative refinement. Evaluating across six diverse benchmarks, we demonstrate that: (1) stable model semantics allow LLMs to naturally express default rules and exceptions, outperforming SMT-based alternatives by significant margins on nonmonotonic tasks; (2) iterative self-correction is the primary driver of performance, effectively replacing the need for handcrafted domain knowledge; (3) compact in-context reference guides substantially outperform verbose documentation, revealing a "context rot" phenomenon where excessive context hinders constraint adherence.

Adam Ishay, Joohyung Lee

Arizona State University · Samsung Research

LLM Self-Correction Logical Reasoning LLM-Based Program Synthesis Nonmonotonic Reasoning Answer Set Programming Neuro-Symbolic Reasoning

Apr 29, 2026cs.MA

The Cost of Consensus: Isolated Self-Correction Prevails Over Unguided Homogeneous Multi-Agent Debate

Multi-agent debate, where teams of LLMs iteratively exchange rationales and vote on answers, is widely deployed under the assumption that peer review filters hallucinations. Yet the failure dynamics of homogeneous debate remain poorly understood, therefore we report findings from a controlled empirical study of teams of

N{=}10

homogeneous agents (Qwen2.5-7B, Llama-3.1-8B, Ministral-3-8B) across

R{=}3

debate rounds on two high-difficulty benchmarks (GSM-Hard and MMLU-Hard). We compare peer debate against isolated self-correction and a stochastic noise control that injects rationales from unrelated problems. We decompose debate failure into three model-dependent pathways: sycophantic conformity, where agents uncritically adopt majority answers (modal adoption up to 85.5%); contextual fragility, where peer rationales destabilize previously correct reasoning (vulnerability rate up to 70.0%); and consensus collapse, where plurality voting discards correct answers already present in the generation pool (oracle gap up to 32.3 percentage points). Ablations over communication density (

K \in \{2,4,9\}

) and sampling temperature (

T \in \{0.4, 0.7\}

) show that conformity reaches high levels at minimal peer exposure (

K{=}2

) and intensifies with greater initial diversity. Across all configurations, debate consumes 2.1-3.4

\times

more tokens (up to 28,631 tokens per problem) than self-correction for equal or lower accuracy. Our results indicate that, within the 7-8B parameter class, homogeneous teams without structured roles do not benefit from unguided peer exchange, and that isolated self-correction consistently offers a more favorable cost-accuracy tradeoff.

Blaž Bertalanič, Carolina Fortuna

Jožef Stefan Institute Ljubljana, Slovenia

LLM Self-Correction AI Agent Evaluation Multi-Agent Debate Multi-Agent Consensus

Apr 28, 2026cs.CV

OmniVTG: A Large-Scale Dataset and Training Paradigm for Open-World Video Temporal Grounding

Video Temporal Grounding (VTG), the task of localizing video segments from text queries, struggles in open-world settings due to limited dataset scale and semantic diversity, causing performance gaps between common and rare concepts. To overcome these limitations, we introduce OmniVTG, a new large-scale dataset for open-world VTG, coupled with a Self-Correction Chain-of-Thought (CoT) training paradigm designed to enhance the grounding capabilities of Multimodal Large Language Models (MLLMs). Our OmniVTG is constructed via a novel Semantic Coverage Iterative Expansion pipeline, which first identifies gaps in the vocabulary of existing datasets and collects videos that are highly likely to contain these target concepts. For high-quality annotation, we leverage the insight that modern MLLMs excel at dense captioning more than direct grounding and design a caption-centric data engine to prompt MLLMs to generate dense, timestamped descriptions. Beyond the dataset, we observe that simple supervised finetuning (SFT) is insufficient, as a performance gap between rare and common concepts still persists. We find that MLLMs' video understanding ability significantly surpasses their direct grounding ability. Based on this, we propose a Self-Correction Chain-of-Thought (CoT) training paradigm. We train the MLLM to first predict, then use its understanding capabilities to reflect on and refine its own predictions. This capability is instilled via a three-stage pipeline of SFT, CoT finetuning, and reinforcement learning. Extensive experiments show our approach not only excels at open-world grounding in our OmniVTG dataset but also achieves state-of-the-art zero-shot performance on four existing VTG benchmarks. Code is available at https://github.com/oceanflowlab/OmniVTG.

Minghang Zheng, Zihao Yin, Yi Yang +2

Wangxuan Institute of Computer Technology, Peking University · Central Media Technology Institute, Huawei Technologies Ltd. · State Key Laboratory of General Artificial Intelligence, Peking University +1

Temporal Video Grounding LLM Self-Correction LLM Grounding

Apr 27, 2026cs.CL

Faithful Autoformalization via Roundtrip Verification and Repair

When an LLM formalizes natural language, how do we know the output is faithful? We propose a roundtrip verification approach which does not require ground-truth annotations: formalize a statement, translate the result back to natural language, re-formalize, and use a formal tool to check logical equivalence. When the two formalizations agree, this provides evidence of a faithful formalization. When they disagree, a stage-level diagnosis localizes the error to a specific translation step, and a scoped repair operator attempts to correct that step. We evaluate the framework on two statutory domains (the Texas Transportation Code and the Texas Parks and Wildlife Code) using two LLMs (Claude Opus~4.6 and GPT-5.2) with three repair baselines. Diagnosis-guided scoped repair is the most effective method, with effectiveness contingent on the reliability of the diagnosis function. Across both domains and both models, under our full repair system, rules that fail the equivalence check show 1.4x-2.5x more natural language inference (NLI) drift than rules that pass it.

Daneshvar Amrollahi, Jerry Lopez, Clark Barrett

Stanford University, USA

LLM Self-Correction Legal NLP Formal Verification Autoformalization

Apr 25, 2026cs.AI

When Corrective Hints Hurt: Prompt Design in Reasoner-Guided Repair of LLM Overcaution on Entailed Negations under OWL2DL

We report a reproducible error pattern in GPT-5.4 on OWL2DL compliance queries: the model frequently answers unknown'' when the reasoner-entailed answer is no'' under \emph{FunctionalProperty} closure or class \emph{disjointness}. Using 180 reasoner-audited queries from a procedural expansion of the observed pattern plus 18 hand-authored held-out queries in two unrelated domains (insurance and clinical), we compare four interaction modes under matched query budget: single-shot, three rounds of generic ``you-are-wrong'' retry, three rounds of reasoner-verdict repair with an open-world-assumption (OWA) hint, and the same repair without the hint. Direct faithfulness is 43.9,% (Wilson 95,% CI

[36.8,51.2]

); generic retry reaches 81.7,% (

[75.4,86.6]

); the verdict-with-hint variant is \emph{worse} at 67.2,% (

[60.1,73.7]

); the verdict-only variant reaches 97.8,% (

[94.4,99.1]

). All pairwise comparisons remain significant under McNemar's exact test with Bonferroni correction (

α= 0.01

; all

p < 10^{-5}

). The same fingerprint accounts for 4/4 errors on the held-out queries. Our interpretation is bounded: prompt framing can matter more than corrective content, and reasoner-guided wrappers should be ablated explicitly.

Yijiashun Qi, Xiang Xu, Yuxuan Li

University of Michigan · ByteDance Inc. · University of Pennsylvania

LLM Self-Correction Logical Reasoning LLM Prompting LLM Reasoning

Apr 24, 2026cs.AI

Self-Correction as Feedback Control: Error Dynamics, Stability Thresholds, and Prompt Interventions in LLMs

Iterative self-correction is increasingly deployed in agentic LLM systems, yet whether repeated refinement improves or degrades performance remains inconsistent across models. We recast self-correction as a closed-loop feedback-control problem in which the same model is both controller and plant, and analyze its error dynamics via a two-state Markov model over {Correct, Incorrect}, parameterized by the Error Introduction Rate (EIR) and Error Correction Rate (ECR). The model yields a directly measurable stability threshold -- iterate only when ECR/EIR > Acc/(1-Acc) -- in which EIR acts as a stability margin and prompting becomes lightweight controller design. Empirically, across 7 models and 3 datasets (GSM8K, MATH, StrategyQA), a sharp near-zero EIR boundary (< 0.5%) cleanly separates beneficial from harmful self-correction: only o3-mini (+3.4 pp), Claude Opus 4.6 (+0.6 pp), and o4-mini (+/-0 pp) stay non-degrading, while GPT-5 and four others lose accuracy. A verify-first prompt intervention then provides causal evidence: it drives GPT-4o-mini's EIR from 2% to 0% and converts a -6.2 pp degradation into +0.2 pp (paired McNemar, p<10^{-4}), with negligible change on already-sub-threshold models -- exactly as the diagnostic predicts. A complementary analysis of adaptive self-consistency (ASC) shows it halts harmful refinement at a 3.8 pp confidence-elicitation cost, exposing a two-tier capability structure: prompt-level EIR suppression prevents degradation, whereas ECR enhancement -- plausibly training-level -- is required for genuine gains. Self-correction should thus be treated not as a default behavior but as a control decision governed by measurable error dynamics.

Aofan Liu, Jingxiang Meng

1Peking University. · University of Chicago.

LLM Self-Correction Language Model-Based Control Causal Interventions in Language Models

Apr 24, 2026cs.LG

How LLMs Detect and Correct Their Own Errors: The Role of Internal Confidence Signals

Large language models can detect their own errors and sometimes correct them without external feedback, but the underlying mechanisms remain unknown. We investigate this through the lens of second-order models of confidence from decision neuroscience. In a first-order system, confidence derives from the generation signal itself and is therefore maximal for the chosen response, precluding error detection. Second-order models posit a partially independent evaluative signal that can disagree with the committed response, providing the basis for error detection. Kumaran et al. (2026) showed that LLMs cache a confidence representation at a token immediately following the answer (i.e. post-answer newline: PANL) -- that causally drives verbal confidence and dissociates from log-probabilities. Here we test whether this PANL signal extends beyond confidence to support error detection and self-correction. Here we test whether this signal supports error detection and self-correction, deriving predictions from the second-order framework. Using a verify-then-correct paradigm, we show that: (i) verbal confidence predicts error detection far beyond token log-probabilities, ruling out a first-order account; (ii) PANL activations predict error detection beyond verbal confidence itself; and (iii) PANL predicts which errors the model can correct -- where all behavioural signals fail. Causal interventions confirm that PANL signals rescue error detection behavior when answer information is corrupted. All findings replicate across models (Gemma 3 27B and Qwen 2.5 7B) and tasks (TriviaQA and MNLI). These results reveal that LLMs naturally implement a second-order confidence architecture whose internal evaluative signal encodes not only whether an answer is likely wrong but whether the model has the knowledge to fix it.

Dharshan Kumaran, Viorica Patraucean, Simon Osindero +2

Google DeepMind · Princeton University

LLM Self-Correction Language Model Error Detection Confidence Estimation in Language Models LLM Interpretability LLM Uncertainty Estimation Metacognition in Language Models

Apr 21, 2026cs.CL

Talking to a Know-It-All GPT or a Second-Guesser Claude? How Repair reveals unreliable Multi-Turn Behavior in LLMs

Repair, an important resource for resolving trouble in human-human conversation, remains underexplored in human-LLM interaction. In this study, we investigate how LLMs engage in the interactive process of repair in multi-turn dialogues around solvable and unsolvable math questions. We examine whether models initiate repair themselves and how they respond to user-initiated repair. Our results show strong differences across models: reactions range from being almost completely resistant to (appropriate) repair attempts to being highly susceptible and easily manipulated. We further demonstrate that once conversations extend beyond a single turn, model behavior becomes more distinctive and less predictable across systems. Overall, our findings indicate that each tested LLM exhibits its own characteristic form of unreliability in the context of repair.

Clara Lachenmaier, Hannah Bultmann, Sina Zarrieß

Computational Linguistics, Department of Linguistics Bielefeld University, Germany

LLM Self-Correction Human-AI Interaction LLM Reliability Conversational Agents

Apr 21, 2026cs.LG

Decompose, Structure, and Repair: A Neuro-Symbolic Framework for Autoformalization via Operator Trees

Statement autoformalization acts as a critical bridge between human mathematics and formal mathematics by translating natural language problems into formal language. While prior works have focused on data synthesis and diverse training paradigms to optimize end-to-end Large Language Models (LLMs), they typically treat formal code as flat sequences, neglecting the hierarchical logic inherent in mathematical statements. In this work, we introduce Decompose, Structure, and Repair (DSR), a neuro-symbolic framework that restructures autoformalization into a modular pipeline. DSR decomposes statements into logical components and maps them to structured operator trees, leveraging this topological blueprint to precisely localize and repair errors via sub-tree refinement. Furthermore, we introduce PRIME, a benchmark of 156 undergraduate and graduate-level theorems selected from canonical textbooks and expertly annotated in Lean 4. Experimental results demonstrate that DSR establishes a new state-of-the-art, consistently outperforming baselines under equivalent computational budgets. The datasets, model, and code are available at https://github.com/XiaoyangLiu-sjtu/DSR.

Xiaoyang Liu, Zineng Dong, Yifan Bai +3

School of Mathematical Sciences, Shanghai Jiao Tong University · Zhiyuan College, Shanghai Jiao Tong University · Institute of Natural Sciences, MOE-LSC, CMA-Shanghai, Shanghai Jiao Tong University

LLM Self-Correction Neuro-Symbolic Reasoning Autoformalization LLM Mathematical Reasoning