ModalityDance/slpo-coconut-gpt2
Text Generation • 0.1B • Updated • 49
Checkpoints for the paper SLPO: Scaling Latent Reasoning via a Surrogate Policy (arXiv:2607.19691).
Note COCONUT+SLPO (GPT-2); gate thr=0.6
Note CODI+SLPO (GPT-2); gate thr=0.7