zjunlp/WKM

[NeurIPS 2024] Agent Planning with World Knowledge Model

38
/ 100
Emerging

This project helps improve how AI agents make decisions in complex virtual environments. It takes data from both ideal and problematic agent behaviors, then processes this information to build a 'world knowledge' model. This model then guides the AI agent to plan better and avoid common errors like blindly trying things or generating nonsensical actions, benefiting researchers and developers working on intelligent agents for simulated tasks.

164 stars. No commits in the last 6 months.

Use this if you are developing AI agents for interactive tasks in simulated environments and want them to plan more effectively and reduce errors by incorporating a deeper understanding of the 'world' they operate in.

Not ideal if you are looking for a pre-trained, ready-to-deploy agent for real-world physical tasks or if you don't have existing trajectory data to train the world knowledge model.

AI agents Reinforcement learning Autonomous systems Cognitive modeling Simulated environments
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 12 / 25

How are scores calculated?

Stars

164

Forks

12

Language

Python

License

Apache-2.0

Last pushed

Dec 17, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/zjunlp/WKM"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.