Nick Cammarata: late-layer receptive fields span whole images, dozens of validation tools exist
nickcammarata · x · 2026-10-22
Liv Gorton argues a neuron explanation is insufficient if, after cropping to its receptive field, humans can't beat chance on predicting activations. Nick Cammarata responds that late-layer receptive fields cover the whole image, making that test just "guess from liked dataset images"—far too weak, and dozens of better tools exist; he's running more experiments.
Related event: Debate Flares Over How to Validate Neuron Interpretations(3 posts)→
More from Research
- EMNLP paper: diffusion LMs internally encode latent denoising timestep representations — gsarti_ · 2026-10-23
- KU Leuven and VIB open two Group Leader professorships in AI for Biology, start 2027 — jpillowtime · 2026-10-23
- Subspace Tuning curbs LLM catastrophic forgetting by freezing principal singular directions — scottniekum · 2026-10-23
- Perceptic launches PercepticOS, an AI workspace for 'scientific superintelligence' — nathanbenaich · 2026-10-23
- EPFL's Bosselut to talk long context as test-time learning at ELLIS seminar — andre_t_martins · 2026-10-23
- DRATS: Adaptive Task Sampling Boosts Worst-Task Performance in Multi-Task RL — Cohere · 2026-10-23