ProActor introduces an RL framework for proactive task scheduling that uses automated time-window annotation, proactiveness metrics, GRPO optimization, and composite rewards to improve timing while preserving action alignment.
Trajectory saving via W&B Weave (Weights & Biases, 2025a) is made optional to mitigate potential sensitive data leakage during training
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ProActor: Timing-Aware Reinforcement Learning for Proactive Task Scheduling Agents
ProActor introduces an RL framework for proactive task scheduling that uses automated time-window annotation, proactiveness metrics, GRPO optimization, and composite rewards to improve timing while preserving action alignment.