BREAKING
Hypothetical Minds agent gains attention
How the ToM module works
1
Form hypotheses
↓
2
Score by accuracy
↓
3
Refine plans
Beyond ReAct and PlanReAct
Baselines
Limits
●
ReAct, PlanReAct
●
Weak high-level planning
Hypothetical Minds
New
●
Perception and memory
●
Two-level planning with ToM
Tested on Melting Pot benchmark
0
GitHub stars
A blueprint for inter-agent tasks
AI NEWS BLITZ
An LLM agent with a Theory of Mind module is drawing renewed attention on GitHub.