Search papers, labs, and topics across Lattice.
2
0
6
3
Folding multi-branch redundancy slashes speculative decoding costs in diffusion LLMs, unlocking up to 2x faster generation.
AgentDoG 1.5 proves you can achieve GPT-5.4-level agent safety with open-source models trained on just 1k samples, slashing deployment overhead by two orders of magnitude.
