New CCN poster finds LLM-brain alignment scales differently across cortical systems
neuranna · x · 2026-08-04
A CCN 2026 poster reports that LLM-brain alignment scales differently across cortical systems.
- In naturalistic story-listening fMRI, encoding performance in the language network improves with representational dimensionality, training progression, model size, and more training stories.
- Category-selective visual regions show scaling patterns similar to the language network.
- Auditory cortex behaves differently: gains are weak or non-monotonic.
- The authors say low-level cues such as word/phoneme rate explain most auditory-cortex performance, while other regions rely on different signals.
- Conclusion: strong LLM-based brain prediction can come from different underlying features depending on the cortical system.
More from Research
- Paper maps 41 agent failure modes to model, harness, tools and environment — omarsar0 · 2026-08-04
- Claude thread proposes an adversarial review skill for agentic science — ChrisGPotts · 2026-08-04
- Singular learning theory talk links neural-network generalization to alignment — andrewgwils · 2026-08-04
- AI paper argues best-of-K boosts generative expressivity, not just sampling quality — anshulkundaje · 2026-08-04
- ASCII art may be a better taste benchmark for frontier models than you think — weswinder · 2026-08-04
- A curated reading list for DeltaNet, FlashKDA, vLLM serving and MoE — austinvhuang · 2026-08-04