MC-EvoBench: Auditing Experience-Driven Self-Improvement in Minecraft Agents
MC-EvoBench tests whether Minecraft agents convert experience into durable capabilities that transfer across reset worlds.
Stories per day · 30 days
2 RSS
Relatedbenchmark1communication1embodied-agents1evolutionary-robotics1mc-evobench1minecraft1multi-agent1self-improvement1sim-to-sim1simulation1
3 stories, newest first · one card per story
Stories tagged #transfer
Evaluating the Transfer of Co-Evolved Communication from 2D to 3D Simulation
Co-evolved robot communication transfers from 2D to 3D physics, but short episodes underestimate success.
EVOKE: Eliciting World Knowledge in Agents for Transferable Decision-Making
EVOKE uses goal diversity to elicit LLM agents' pretrained world knowledge for better transfer.