Search papers, labs, and topics across Lattice.
Affiliation:
3
0
4
2
Batching future MH proposals unlocks significant speedups for LLM power sampling without altering the core reasoning process.
A single controller can now robustly learn stochastic transport across a range of initial distribution shifts, outperforming standard SB and noise augmentation on challenging tasks.
A novel inference-time sampling strategy unlocks frontier LLM reasoning capabilities in smaller models by resolving the exploration-exploitation dilemma.
