Research map / Vision-language and multimodal / Human and embodied modeling
Disentangling the Factors of Convergence between Brains and Computer Vision Models
Joséphine Raugel, Marc Szafraniec, Huy V. Vo, Camille Couprie, Jérémy Rapin, Stéphane d'Ascoli, Patrick Labatut, Piotr Bojanowski, Valentin Wyart, Jean-Remi King
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| increase model size from Small to Giant (trained from scratch) (ablation) | DINOv3 (human-centric 1.7B images) · NSD fMRI/THINGS-MEG, end of training | encoding score (Pearson R): 0.096 -> 0.107 | Improved |
| train DINOv3 on human-centric images instead of satellite or cellular images (10M images each) (ablation) | DINOv3 Large · NSD fMRI and THINGS-MEG | encoding, spatial and temporal scores | Improved |
| train DINOv3 longer, from untrained to trained (ablation) | DINOv3 7B · fMRI average encoding score | encoding score (Pearson R): .03 -> .09 | Improved |
| train on non-human-centric data (satellite, cellular) (ablation) | DINOv3 Large · brain similarity across image types | encoding score | Improved |
| increase model size (ablation) | DINOv3 · higher-level cortices (BA44, IFS) vs early visual (V1, V2) | encoding score | Mixed |
| compare half time to anatomical distance from V1 | DINOv3 7B training · fMRI ROIs | correlation: R=0.91 | Improved |
Closest papers by meaning
- One Hundred Neural Networks and Brains Watching Videos: Lessons from Alignment (ICLR 2025 · similarity 0.91)
- Bridging Critical Gaps in Convergent Learning: How Representational Alignment Evolves Across Layers, Training, and Distribution Shifts (NeurIPS 2025 · similarity 0.91)
- Dimensionality Mismatch Between Brains and Artificial Neural Networks (NeurIPS 2025 · similarity 0.91)
- Evaluating alignment between humans and neural network representations in image-based learning tasks (NeurIPS 2024 · similarity 0.89)
- Characterizing Universal Object Representations Across Vision Models (NeurIPS 2026 · similarity 0.88)
- A Convolutional Neural Network Interpretable Framework for Human Ventral Visual Pathway Representation (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
