NewA calmer, faster ZeroHour — security assessments on articles, story timelines and search from any page.

#soft-mtp

0 stories this week · 1 this month · 1 in 6 months · mostly AI

Stories per day · 30 days

Relatedlanguage-model-training1 multi-token-prediction1 speculative-decoding1

1 stories, newest first · one card per story

Hugging Face daily papers15d ago

Soft-MTP drops a noisy residual in multi-token prediction training

From 167M to 1.2B parameters, Soft-MTP lowers next-token perplexity and speeds speculative decoding without new parameters.

multi-token-prediction language-model-training