New preprint derives representational geometry whose distances equal mutual information
ShahabBakht · x · 2026-11-18
A new preprint argues that representational geometry depends on building assumptions, and derives one from a few simple requirements. The resulting distances turn out to be exactly tied to mutual information, contrasting Fisher geometry (fine discrimination) with multi-scale geometry (mutual information) in the neural-code → geometry → performance pipeline.
More from Research
- Research agenda: 'self-inoculation' to keep models' minds intact through RL training — repligate · 2026-11-18
- Embed team shows 3 years of fMRI-to-image BCI work, betting brain interfaces help align superintelligence — jonxuxu · 2026-11-18
- Robo-ValueRL rates actions by estimated task progress to learn from messy robot data — burkov · 2026-11-18
- RFD4-Proteina protein design GUI now runs in your browser, even on phones — AllThingsApx · 2026-11-18
- DeepBA: a 2M-parameter GNN learns bundle adjustment and beats prior learned SfM from a cold start — Singularitarian · 2026-11-18
- Redditor Asks: Can AI Now Formalize All of Math, CS, and Theoretical Physics in Lean? — SignificanceBulky162 · 2026-11-18