Google DeepMind and UC Berkeley propose HOPE, a framework for progressively decomposing network weights
burkov · x · 2026-07-25
In a new paper from Google DeepMind and UC Berkeley, researchers introduce Hilbert Operator for Progressive Encoding (HOPE), a mathematical framework that gradually deconstructs representations embedded in trained network weights.
The post does not go into the full derivation, but points readers to the paper and suggests reading it with an AI tutor because the material is dense.
Related event: DeepMind and Berkeley Propose HOPE to Quantify Neuron Capacity(5 posts)→
More from Research
- Google DeepMind pilots world's first double-blind evaluations for frontier models — scychan_brains · 2026-10-26
- Rewriting scaling law study: optimal synthetic data share rises from 10% to 32% with compute — heghbalz · 2026-10-26
- FLASH: GPU-native deformable sim learns cloth folding in minutes on one RTX 5090 — ChongZzZhang · 2026-10-26
- LEMON: zero-shot LLM misbehavior monitor flags hallucinations from attention maps — yoavartzi · 2026-10-26
- Behind the EMNLP keynote: LMs as simulated learners to generate child language acquisition hypotheses — najoungkim · 2026-10-26
- The paper behind the EMNLP 2026 keynote on bigram 'simulated learners' — najoungkim · 2026-10-26