Nemotron-Labs-Sigma Collection Continuous diffusion language models (3B/8B): base, instruct, few-step distilled (NFE 2/4/8/16) and self-speculative variants. • 14 items • Updated about 22 hours ago • 7
Think Less, Locally: Local Preference Learning for Reasoning Compression Paper • 2611.04814 • Published 15 days ago • 67
Tree-Ising: Repairing Language Models with Bit-Level Ising Edits Paper • 2611.14210 • Published 8 days ago • 46
Test-Time Scaling for Diffusion Language Models with Frequency-Aware Remasking Paper • 2611.10486 • Published 11 days ago • 9
The Surprising Effectiveness of Deleting Weights in LLM Post-Training Paper • 2611.10866 • Published 11 days ago • 22
Running 283 The ultimate guide to RL environments: building and scaling them in the LLM era 📝 283 Building and scaling RL environments for LLM training