bycloud

A very valuable log of a high-level research program A great idea is only the first step…

RL is no longer needed? "Neural Thickets: Diverse Task Experts Are Dense Around Pretrained Weights" This paper argues that large pretrained models don’t sit at a single optimal set of weights but inside a dense “thicket” of nearby task-specific experts. So once pretraining is strong enough, randomly sampling small weight perturbations often yields specialists that outperform the base model on different tasks, and simply selecting and ensembling these guesses (RandOpt) can rival standard post-training methods. This suggests that much of what post-training does is just selecting useful behaviors already latent around the pretrained weights rather than learning entirely new ones.

Haven't gotten around to writing in a bit, here's a short blog on my thoughts since releasing RLMs on the state of AI research. A stronger belief I hold is that future LMs will be scaffolds, and that current LMs are already far more capable than we use them for!