Awesome Papers Autonomous Agent

Report Abuse

Basic Information

This repository is a curated, actively maintained collection of recent research papers and resources about autonomous agents. It focuses on two main types of agents called out in the README: RL-based agents and LLM-based agents, and it organizes literature into topical sections such as instruction following, world models, language-as-knowledge, LLMs as tools, multimodal agents, multi-agent systems, benchmarks and datasets, algorithm design, continual learning and applications. The README states a scope that emphasizes recent developments rather than traditional RL agents and lists surveys, conference papers, project pages and links to code when available. The project includes a table of contents, update history and classification criteria to help readers quickly find papers relevant to planning, benchmarking, multimodality, and combinations of reinforcement learning and large language models. The collection invites issues and contributions for missing papers and is intended to track the evolving autonomous agent literature.

Links

Categorization

App Details

Features
Organized bibliography with a clear table of contents that separates RL-based and LLM-based agent work. Subsections cover instruction following, world models, language-as-knowledge, LLM-as-tool, generalization, continual learning, transformer-based policies, trajectory-to-language, multimodal agents, task-specific designs, multi-agent collaboration, experimental analyses, benchmarks and datasets, algorithmic design, and combined RL/LLM approaches. Entries often include paper titles, venue tags and links to arXiv, project pages or GitHub repos when available. The README documents update history and maintenance notes indicating active curation. The repo highlights surveys and key recent conference papers and groups related resources (code, projects, blogs) next to entries when provided in the original sources.
Use Cases
This collection helps researchers, students and practitioners rapidly survey the autonomous agent literature by topic and research trend. Users can find surveys for overviews, landmark and recent conference papers for deep dives, and pointers to project pages or code for reproducibility and experiments. The topical organization makes it easier to prepare literature reviews, identify benchmarks and datasets, compare algorithmic approaches and discover work that combines LLMs with reinforcement learning or explores multimodal and multi-agent settings. The active maintenance and update log let users rely on the list to stay current, and links to repositories or project pages facilitate following up on implementation details and evaluations without searching individual venues manually.

Please fill the required fields*