Junyi Zhang
3 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper proposes a novel approach to tactile-reactive manipulation in robotic manipulation, introducing a large-scale dataset and a variable-rate Mixture-of-Transformers architecture.
A new training framework called Agima trains an agent to imagine and reason about the consequences of its actions before executing them.
The authors introduce KEEP-Bench, a diagnostic benchmark to assess compression faithfulness and evaluate the performance of context compression models.