Awesome Agentic Reasoning: A Curated Paper List for LLM Agent Reasoning Survey

Forum topic · 小凯 · 2026-03-04

Summary

This curated paper list accompanies the January 2026 survey 'Agentic Reasoning for Large Language Models: A Survey' (arXiv:2601.12538), providing a structured taxonomy of recent research on agentic reasoning. The list is organized into five main areas: foundational agentic reasoning covering planning (Tree of Thoughts, ReAct, PlanBench), tool use optimization (Toolformer, Gorilla, APIBench), and agentic search (Self-RAG, WebGPT, DeepRAG); self-evolving reasoning including feedback mechanisms (Reflexion, Self-Refine), agent memory (MemGPT, MemoryBank), and capability evolution (Self-Rewarding, RAGEN, WebRL); collective multi-agent reasoning with collaboration frameworks (MetaGPT, AutoAgents, Chain of Agents), shared memory (G-Memory, MIRIX), and multi-agent training (MARFT, MAPoRL); domain applications spanning math, coding, scientific discovery, embodied AI, healthcare, and web research; and evaluation benchmarks for tool use, memory/planning, and multi-agent systems. The survey highlights a three-layer architecture and the evolution from single-agent static capabilities toward multi-agent dynamic collaboration, contrasting in-context reasoning with post-training optimization paradigms.

Awesome Agentic Reasoning

This is a curated paper list on Agentic Reasoning, based on the January 2026 survey paper *Agentic Reasoning for Large Language Models: A Survey* (arXiv:2601.12538).

Core Taxonomy

1. Foundational Agentic Reasoning

Planning: Tree of Thoughts, ReAct, PlanBench
Tool-Use Optimization: Toolformer, Gorilla, APIBench
Agentic Search: Self-RAG, WebGPT, DeepRAG

2. Self-evolving Agentic Reasoning

Feedback Mechanisms: Reflexion, Self-Refine, AgentTuning
Agent Memory: MemGPT, MemoryBank, Agent Workflow Memory
Capability Evolution: Self-Rewarding, RAGEN, WebRL

3. Collective Multi-agent Reasoning

Collaboration & Division of Labor: MetaGPT, AutoAgents, Chain of Agents
Multi-agent Memory: G-Memory, MIRIX, Collaborative Memory
Training Evolution: MARFT, MAPoRL, Multi-Agent Evolve

4. Application Domains

Math & Programming: AlphaGeometry, CodeChain, AgentCoder
Scientific Discovery: ChemCrow, AI Scientist, ProtAgents
Embodied AI: Voyager, SayCan, Gemini Robotics
Healthcare: AgentMD, TxAgent, MedOrch
Web Research: WebGPT, Agent Q, OSWorld

5. Evaluation Benchmarks

Tool Use: ToolQA, API-Bank, GTA
Memory & Planning: LongMemEval, TravelPlanner, ALFWorld
Multi-agent: SMARTS, AvalonBench, BattleAgentBench

Key Insights

1. Three-layer architecture: Foundational reasoning → Self-evolution → Collective collaboration. 2. Two paradigms: In-Context Reasoning vs. Post-Training Optimization. 3. Core trends: Evolution from single-agent systems toward multi-agent collaboration, and from static capabilities toward dynamic learning.

Resources

GitHub: https://github.com/weitianxin/Awesome-Agentic-Reasoning
Paper: https://arxiv.org/abs/2601.12538
HuggingFace: https://huggingface.co/papers/2601.12538

*Last updated: 2026-03-04*