Search papers, labs, and topics across Lattice.
Shanghai Jiao Tong University
2
0
4
LLMs waste compute on tokens that have already "figured it out" – DASH selectively skips these tokens during prefill, speeding things up without retraining or sacrificing accuracy.
Forget freezing your feature extractor: TALON unlocks on-the-fly category discovery by continuously learning from unlabeled data during test time, outperforming fixed-knowledge methods.