conftrace_

token-by-token policy

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

hierarchical multi-turn rl (1) off-policy rl (1) actor-critic framework (1) delayed reward (1) language model agent (1)

Papers

ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RL ICML 2024