Claude Code learns from your corrections: self-correcting memory that compounds over 50+ sessions. Context engineering, parallel worktrees, agent teams, and 17 battle-tested skills.

Enhancing LLMs with LoRA

🔥 [CVPR2024] Official implementation of "Self-correcting LLM-controlled Diffusion Models (SLD)

Explore concepts like Self-Correct, Self-Refine, Self-Improve, Self-Contradict, Self-Play, and Self-Knowledge, alongside o1-like reasoning elevation🍓 and hallucination alleviation🍄.

Cognitive prisms for structural analysis — any domain, any model. 1000+ experiments, 58 prisms, 27 scan modes. The cheapest model + right prism beats the most expensive without one. Cross-architecture validated: Claude + Gemini converge on the same structural laws.

[ICLR 2025] SuperCorrect: Advancing Small LLM Reasoning with Thought Template Distillation and Self-Correction

List of papers on Self-Correction of LLMs.

A comrephensive collection of learning from rewards in the post-training and test-time scaling of LLMs, with a focus on both reward models and learning strategies across training, inference, and post-inference stages.

Introduces a comprehensive agent-based framework for guided software development (GSD)

Scalable long read self-correction and assembly polishing with multiple sequence alignment

[ACL'24] Code and data of paper "When is Tree Search Useful for LLM Planning? It Depends on the Discriminator"

LoopCraft runtime harness plugin for Hermes with evidence-first execution, verification-first workflows, and self-correcting AI agent behavior. Installs as proofrail for compatibility.

[ICLR 2025] Breaking Mental Set to Improve Reasoning through Diverse Multi-Agent Debate

SCoRe: Training Language Models to Self-Correct via Reinforcement Learning

[ICLR 2025 SSI-FM] Self-Taught Self-Correction for Small Language Models

Evolving agent harnesses: a research program on how far N orchestrated calls of a small model can rival a frontier model. We evolve the harness (structure + prompts) with reflective optimizers + a verified-acceptance gate.

SAGE: self-correcting autonomous research via Multi-Hypothesis Failure Attribution (MHFA). Diagnose many causes of a failed experiment, verify the most critical, and route the fix to the right level.

Self-correction engine for AI coding agents - implements the Reflexion paper (Shinn et al., 2023) as a Rust MCP server. Turns failures into persistent, searchable lessons that prevent the same mistakes across sessions.

Official Implementation of "ProgCo: Program Helps Self-Correction of Large Language Models" (Accpeted at ACL2025 Main)

Residual-stream confabulation detection and self-correction for language models.