AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware SSL

Sangyoon Bae, Mehdi Azabou, Blake Aaron Richards, Jiook Cha

ICLR 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Data-efficient and robust self-supervision

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 changedWhereMeasuredEffect
pretrain on predictable neurons then finetune on unpredictablefrom-scratch POYO+ · movie decodingSSIM: 0.528 -> 0.593Improved
pretrain on predictable neurons then finetune on unpredictablefrom-scratch POYO+ · drifting gratingsaccuracy: 0.492 -> 0.555Improved
reverse curriculum: pretrain on unpredictable neurons (ablation)POYO-CAP · drifting gratingsaccuracy: 0.555 -> 0.213Hurt
select neurons by statistical regularity instead of random subset (ablation)POYO-CAP · drifting gratingsaccuracy: 0.254 -> 0.555Improved
select predictable neurons instead of mixed with unpredictable (ablation)POYO-CAP · movie decodingSSIM: 0.543 -> 0.593Improved
use all predictable lines instead of inhibitory only (ablation)POYO-CAP · movie decodingSSIM: 0.544 -> 0.593Improved
temporal masking instead of random masking (ablation)POYO-CAP · movie decodingSSIM: 0.540 -> 0.593Improved
add auxiliary cross-entropy loss to masked reconstruction (ablation)POYO-CAP · drifting gratingsaccuracy: 0.099 -> 0.555Improved
add masked reconstruction to cross-entropy (ablation)POYO-CAP · drifting gratingsaccuracy: 0.452 -> 0.555Improved
CE weight 0.01 versus 0.1 (ablation)POYO-CAP · movie decodingSSIM: 0.552 -> 0.593Improved
add UNet skip connections (ablation)POYO+ encoder with UNet decoder · movie decodingSSIM: 0.466 -> 0.593Improved
pretrain on predictable neurons onlymodel scaling · SSIM vs model sizescaling slope: slope 0.018 vs approx 0.005-0.013Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy