Tag
This paper proposes a four-layer architecture for cognitive digital twins that enables self-evolving operational loops through integrated cognitive capabilities and task-oriented feedback mechanisms.
This paper proposes TONIC, a token-centric semantic communication framework for task-oriented wireless systems that assigns utility-aware unequal error protection to tokens and uses confidence-aware gating with a Transformer-based completion model, outperforming baselines on image classification.
DexJoCo introduces a benchmark and toolkit for task-oriented dexterous manipulation in MuJoCo, featuring 11 functional tasks, a low-cost data collection system, and comprehensive evaluations that highlight limitations in current dexterous manipulation policies.