AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Asymmetric Contrastive Objectives for Efficient Phenotypic Screening

Luke Nightingale, Joseph Tuersley, Scott Warchal, Andrea Cairoli, Jacob Howes, Cameron Shand, Andrew J Powell, Darren V.S. Green, Amy Strange, Michael Howell

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 1 citations (OpenAlex)

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
replace softmax classifier with contrastive loss using class vectors (NTXent) (ablation)WSL (ResNet18) · BBBC021 well mAPmAP: .790 -> .892Improved
normalize vectors and add temperature (L2-softmax) without contrastive form (ablation)WSL (ResNet18) · BBBC021mAP: .790 -> .803No clear effect
remove repulsive terms from class-vector updates (SPC) (ablation)NTXent (ResNet18) · BBBC021 well mAPmAP: .890 -> .924Improved
remove repulsive terms from class-vector updates (ablation)SupContrast (ResNet18) · BBBC021 well mAPmAP: .894 -> .900Improved
confine class vectors to sphere and split optimizers, keeping repulsion (ablation)NTXent (ResNet18) · BBBC021 well mAPmAP: .892 -> .890No clear effect
remove repulsion from model (encoder) update (ablation)cosine loss · BBBC021mAP: well mAP .284Hurt
use SPC instead of WSL (ablation)ResNet18 · HaCaT CRISPR unseen perturbationsmAP: .370 -> .483Improved
use SPC instead of WSL (ablation)ResNet18 · HaCaT JUMP unseen perturbationsmAP: .240 -> .298Improved
use SPC as fine-tuning objective instead of WSL (ablation)DINOv2 ViT/B with LoRA r=16 · HaCaT CRISPR cell holdoutmAPImproved
scale fine-tuned model from ViT/B to ViT/L (ablation)SPC (DINOv2 LoRA) · HaCaT CRISPR cell holdoutmAPImproved
use temperature tau=0.25 instead of 0.1 or 1 (ablation)contrastive losses · BBBC021MoA metricsImproved
use SPC-ResNet18 (11M) instead of Phenom-1 (307M)RxRx3-core · biological recall CORUMrecall: .395 -> .428Mixed

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