Scaled reproduction of arXiv:2601.15165 with H200 Jobs, raw outputs, released checkpoints, Trackio logbook, and poster.
-
The Flexibility Trap: Why Arbitrary Order Limits Reasoning Potential in Diffusion Language Models
Paper • 2601.15165 • Published • 75 -
GSAI-ML/LLaDA-8B-Instruct
Text Generation • 8B • Updated • 444k • 364 -
nzl-thu/LLaDA-Instruct-JustGRPO-GSM8K
Text Generation • 8B • Updated • 132 • 8 -
jgallego9/flexibility-trap-repro-results
Preview • Updated • 38