This repository curates resources on Large Language Models (LLMs) for planning and reasoning, targeting researchers and developers. It provides a comprehensive overview of techniques, limitations, benchmarks, and related papers to advance LLM capabilities in complex, real-world applications.

How It Works

The collection categorizes resources into key areas: Techniques (e.g., Chain-of-Thought, Tree of Thoughts, ReAct), Reasoning Limitations (papers analyzing LLM failures), and Benchmarks (evaluations like AgentBench, PlanBench). This structured approach allows users to quickly identify relevant research and understand the current state of LLM planning and reasoning.

Quick Start & Requirements

This is a curated list of papers and resources, not a runnable codebase. No installation or specific requirements are needed to browse the content.

Highlighted Details

• Features papers on novel techniques like Chain-of-Thought, Tree of Thoughts, and ReAct.
• Includes critical investigations into LLM reasoning limitations and failures.
• Compiles a wide range of benchmarks for evaluating LLM planning performance.
• Provides links to papers, code repositories, and project pages for further exploration.

Maintenance & Community

The repository is maintained by Sambhav Khurana and contributors. Users are encouraged to contribute and cite the associated papers (e.g., arXiv:2502.12521, arXiv:2502.19295).

Licensing & Compatibility

The repository itself is a collection of links and does not have a specific license. Individual papers and code repositories linked within will have their own licenses.

Limitations & Caveats

This repository is a curated list and does not provide executable code or direct access to LLM models. Users must independently access and evaluate the linked resources.