MIT & Adobe's Latent Lattice LMs cut perplexity 15.41→14.40, matching 50% larger Transformers

burny_tech · x · 2026-11-19

Researchers from MIT and Adobe (Shivam Duggal, William Freeman, Antonio Torralba, Richard Zhang, Eli Shechtman, et al.) introduce Latent Lattice Language Models (LLLMs), a new axis for scaling LM computation: increasing the number of active computational states rather than depth or looped layers.

Takeaway: persistent per-layer state is a complementary scaling axis for language-model compute.

Original post →

More from Models

Models channel →