Search papers, labs, and topics across Lattice.
2
0
5
16
Forget runtime inference costs: "compiled AI" slashes token consumption by 57x while maintaining accuracy by pre-generating deterministic code from LLMs.
Uniform-state diffusion models, often overlooked in favor of masked diffusion, surprisingly outperform autoregressive and masked diffusion models on GSM8K when scaled to 1.7B parameters, despite worse perplexity.
