technique

Observe-and-Act Adaptive Context Selection

techniqueactiveprovisionalobserve-and-act-adaptive-context-selection-4b1f312c·1 events·first seen 18d ago

Aliases: Observe-and-Act Adaptive Context Selection

Co-occurring entities

YutongWang1216 3E Memory Module Reinforcement Learning Loong

More like this (12)

Observation-Guided Video-Context Routing In-Context Reward Adaptation Latent Context Language Models language-adaptive switch Contextual Retrieval Learning Red Agent Policy from Observations for Neurosymbolic Autonomous Cyber Agents Observation-Reason-Action Layer-Adaptive Expert Pruning Pose-ICL: 3D-Aware In-Context Learning for Pose-Controllable Subject Customization In-Context Multiple Instance Learning Reinforcement Learning Elicits Contextual Learning of Unseen Language Translation embodiment-aware prompt conditioning

Recent events (1)

5arXiv · cs.CL·18d ago·source ↗

Loong: A Human-Like Long Document Translation Agent with Observe-and-Act Adaptive Context Selection

Loong is a long document translation agent that uses a 3E memory module (Essence-Exemplar-Entity) to store structured historical context, replacing passive full-context attention with RL-optimized adaptive context selection. The agent learns its context retrieval policy via reinforcement learning on self-sampled reasoning trajectories. Evaluations show average gains of up to 13.0 points across three metrics in English↔Chinese, German, and French translation directions, with strong generalization and robustness to noise in ultra-long documents.

Long Context Evolution Agent and Tool Ecosystem YutongWang1216 3E Memory Module Reinforcement Learning +3 more