Loading activity…
@rosinality
ML Engineer
AI
Hinted scaling laws use easier benchmark versions; researchers report median forecasting error fell from 21 to 4.3 percentage points.
AI
A user linked to an arXiv paper they described as fixing the issue; another user praised the work.
AI
A user links an arXiv paper and calls Tencent one of few companies still publishing papers on environment synthesis.
AI
The method reuses top-layer activations in lower layers during inference on existing models.