thu-nics/MARSHAL
MARSHAL: Incentivizing Multi-Agent Reasoning via Self-Play with Strategic LLMs
This project helps AI researchers and developers working on advanced AI agents. It provides a framework for training Large Language Models (LLMs) to perform better in multi-agent environments, like strategic games or complex reasoning tasks. Researchers can input various game environments or reasoning benchmarks to train LLMs that exhibit improved strategic decision-making and generalization capabilities.
Use this if you are an AI researcher or developer building multi-agent systems and want to train LLMs that can strategize effectively in competitive and cooperative scenarios.
Not ideal if you are an end-user looking for a pre-built application or a low-code solution for general AI tasks.
Stars
39
Forks
1
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 13, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/thu-nics/MARSHAL"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
ai4co/reevo
[NeurIPS 2024] ReEvo: Large Language Models as Hyper-Heuristics with Reflective Evolution
SALT-NLP/collaborative-gym
Framework and toolkits for building and evaluating collaborative agents that can work together...
Gen-Verse/LatentMAS
Latent Collaboration in Multi-Agent Systems
lean-dojo/LeanCopilot
LLMs as Copilots for Theorem Proving in Lean
WooooDyy/AgentGym-RL
Code and implementations for the paper "AgentGym-RL: Training LLM Agents for Long-Horizon...