AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Complementary Benefits of Contrastive Learning and Self-Training Under Distribution Shift

Saurabh Garg, Amrith Setlur, Zachary Chase Lipton, Sivaraman Balakrishnan, Virginia Smith, Aditi Raghunathan

NeurIPS 2023 · Label-efficient and robust learning › Semi-supervised learning · approach: Self-training theory · 3 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
Add self-training (FixMatch) on top of contrastive pretraining (STOC vs CL)CL · UDA averageaccuracy: 62.39 -> 67.59Improved
Add contrastive pretraining under self-training (STOC vs ST)ST · UDA averageaccuracy: 56.98 -> 67.59Improved
Add self-training to ERMERM · UDA averageaccuracy: 48.90 -> 56.98Improved
Add contrastive pretraining to ERMERM · UDA averageaccuracy: 48.90 -> 62.39Improved
Self-training vs contrastive learning on Entity13ST vs CL · UDA Entity13accuracy: 77.93 -> 76.58Mixed
Add STOC over CL on VisdaCL · UDA Visda (2 tgts)accuracy: 63.49 -> 70.08Improved
Add STOC over CL on CIFAR->CINICCL · UDA CIFAR->CINICaccuracy: 77.51 -> 79.94Improved
Add self-training over contrastive pretraining in SSLCL · SSL averageaccuracy: 79.27 -> 80.06Improved
Add self-training over CL in SSL on VisdaCL · SSL Visdaaccuracy: 97.67 -> 97.70No clear effect
Contrastive pretraining on source unlabeled data only instead of source+target (ablation)CL · UDA Visda (2 tgts)accuracy: 63.2 -> 33.3Hurt
Self-training over source-only CL vs standalone CL (ablation)STOC (source only) · UDA Living17accuracy: 67.3 -> 75.0Mixed
Add self-training on top of CL on target linear probeCL · UDA linear probeaccuracyNo clear effect

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