First survey of Weight Space Learning: unified taxonomy of understanding, representation, generation
burny_tech · x · 2026-11-10
- New arXiv survey A Survey of Weight Space Learning (arXiv:2603.10090) by Xiaolong Han et al. (11 authors) provides the first unified taxonomy of WSL
- Core claim: neural weights are not just a training end product — weight space itself contains rich structure, with pretrained models forming organized distributions and exhibiting symmetries that can be embedded, compared, or generated
- Three-dimension taxonomy:
- WSU (Understanding): geometry and symmetries of weights
- WSR (Representation): learning embeddings over model weights
- WSG (Generation): synthesizing new weights via hypernetworks or generative models
- Applications: model retrieval, continual and federated learning, NAS, data-free reconstruction
More from Research
- PNAS Nexus paper demonstrates experience-based training in a living tissue model — burny_tech · 2026-11-10
- NVIDIA paper: masking ~10 'intrusion heads' recovers 57% of distractor accuracy loss — rohanpaul_ai · 2026-11-10
- The Five Differential Operators of Physics, All Built From One Symbol: Nabla — burny_tech · 2026-11-10
- Control Limit Hypothesis: High Dopamine in Schizophrenia Is Compensation, Not Cause — burny_tech · 2026-11-10
- Interp Researcher's 2026 Advice: Be Ambitious, Pragmatic, and Go Meta — burny_tech · 2026-11-10
- WorldFoundry: open-source unified infra for evaluating world models — OpenEnvisionLab · 2026-11-10