NeurIPS Oral derives a geometry of stimulus space tied to neural mutual information
ShahabBakht · x · 2026-11-09
A NeurIPS 2026 Oral (shared by simoneazeglio) asks: when a monkey views a frog image, which image features does visual cortex activity preserve, and which does it discard?
- The authors recorded spike activity from neurons in macaque V1 and V4 in response to image stimuli.
- They derive a geometry of stimulus space that is exactly tied to the mutual information carried by the neurons, turning "what does neural coding preserve?" into a precise geometric framework.
- The tweet links to a 12-part thread with an illustration of the experimental setup (monkey viewing images, V1/V4 spike recordings).
More from AGI Musings
- Hot take: good AI judgment is now the new median reviewer — 3scorciav · 2026-11-09
- AI insider warns public perception will worsen as frontier progress outpaces intuition — manosaie · 2026-11-09
- NYU paper: AI assistants push retail investors into index funds by 7.6pp — danielrock · 2026-11-09
- Toby Ord: METR's 50% Time Horizon Can Go to Infinity Without Transforming the World — tobyordoxford · 2026-11-09
- 'Argon is a fun model to think with': an AI's eerie first-person riff on reading and death — aiamblichus · 2026-11-09
- Stanford prof rebuts PhD stipend "injustice" claim: NYC/Bay costs ≠ Moscow/Mumbai — anshulkundaje · 2026-11-09