Tag
本文提出 Discord 与 BRACE 两种方法，利用奖励分歧来校准 MCP 环境中工具使用语言模型的强化学习 rollout 预算分配，并在 Qwen3-VL 多个规模及 BFCL、NESTFUL 基准上验证其优于基线。
This paper introduces PAIR, a pairwise-aware inclusion reweighting method for adaptive rollout allocation in RLVR, improving sample efficiency and accuracy over pointwise allocators by correcting biases in pairwise gradient estimation.
This paper introduces SARA, a sequential adaptive rollout allocation method for RLVR that abandons saturated groups early and reallocates the budget, achieving comparable accuracy with 22% fewer rollouts than dynamic sampling and up to 67% savings when combined.
TRACE is a unified rollout budget allocation framework that enhances reward contrast in multi-turn agentic reinforcement learning by dynamically distributing resources across tree-structured rollouts based on prefix-level informativeness. It improves efficiency and accuracy on agentic benchmarks like Multi-Hop QA.