ICML 2026 accepted papers on Label-efficient and robust learning
352 of the 6,341 papers accepted at ICML 2026 (5.6%) are on Label-efficient and robust learning, compared with 6.8% at ICML 2025. They fall into 9 topics.
See ICML 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Recognition and long-tail · 169 papers
- FLIP2: Expanding Protein Fitness Landscape Benchmarks for Real-World Machine Learning Applications (3 citations)
- Asymmetric Contrastive Objectives for Efficient Phenotypic Screening (1 citation)
- GeoPT: Scaling Physics Simulation via Lifted Geometric Pre-Training (1 citation)
- A Conflict-aware Evidential Framework for Reliable Sleep Stage Classification
- A Refined Generalization Analysis for Extreme Multi-class Supervised Contrastive Representation Learning
- A robust PPG foundation model using multimodal physiological supervision
- Ambiguous Strategic Classification
- An Information-Theoretic Criterion for Efficient Data Synthesis
- Annotations Mitigate Post-Training Mode Collapse
- Beyond Model Base Retrieval: Weaving Knowledge to Master Fine-grained Neural Network Design
- Biased Generalization in Diffusion Models
- BLOCK-EM: Preventing Emergent Misalignment via Latent Blocking
and 157 more in the interactive map.
Self-supervised and contrastive · 45 papers
- CalM: A Self-Supervised Foundation Model for Population Dynamics in Calcium Imaging Data (1 citation)
- A Solvable High-Dimensional Model Where Nonlinear Autoencoders Learn Structure Invisible to PCA While Test Loss Misaligns With Generalization
- AudioMosaic: Contrastive Masked Audio Representation Learning
- BAT: Better Audio Transformer Guided by Convex Gated Probing
- Bayesian Gated Non-Negative Contrastive Learning
- Beyond Instance-Level Self-Supervision in 3D Multi-Modal Medical Imaging
- Bio-Inspired Self-Supervised Learning for Wrist-worn Accelerometer Data
- Brep2Shape: Boundary and Shape Representation Alignment via Self-supervised Transformers
- Directional Neural Collapse Explains Few-Shot Transfer in Self-Supervised Learning
- Evaluating the Representation Space of Diffusion Models via Self-Supervised Principles
- FairSSL: Fair Multimodal Self-Supervised Learning
- Feedback Control for Multi-Objective Graph Self-Supervision
and 33 more in the interactive map.
Zero-/few-shot and meta-learning · 28 papers
- A Semantically Consistent Dataset for Data-Efficient Query-Based Universal Sound Separation
- Alethia: a Foundational Encoder for Voice Deepfakes
- CAOS: Conformal Aggregation of One-Shot Predictors
- Contrastive Geometric Learning Unlocks Unified Structure- and Ligand-Based Drug Design
- Cross-Subject Modeling for Widefield Calcium Imaging via Atlas-Aligned Spatiotemporal Tokenization
- CURVE: Learning Causality-Inspired Invariant Representations for Robust Scene Understanding via Uncertainty-Guided Regularization
- DecoderTCR: Compositional Pretraining and Entropy-Guided Decoding for TCR-pMHC Interactions
- Dynamics Are Learned, Not Told: Semi-Supervised Discovery of Latent Dynamics Geometries For Zero-Shot Policy Adaptation
- Fleet: Few Shots Lead Effective AI-generated Image Detection
- From Zero to Hero: Advancing Zero-Shot Foundation Models for Tabular Outlier Detection
- Minimizing Mismatch Risk: A Prototype-Based Routing Framework for Zero-shot LLM-generated Text Detection
- Model Fusion via Retrofitting
and 16 more in the interactive map.
Domain adaptation and generalization · 28 papers
- A Mechanistic Analysis of Sim-and-Real Co-Training in Generative Robot Policies
- Anti-causal domain generalization: Leveraging unlabeled data
- Bayesian Meta-Learning with Expert Feedback for Task-Shift Adaptation through Causal Embeddings
- Conformal Calibration Transfer
- DADP: Domain Adaptive Diffusion Policy
- Diffract: Spectral View of LLM Domain Adaptation
- Domain Transfer Becomes Identifiable via a Single Alignment
- Expectation Consistency Loss: Rethink Confidence Calibration under Covariate Shift
- Explaining Concept Shift with Interpretable Feature Attribution
- Explaining Data Mixing Scaling Laws
- Exploring Accurate and Transparent Domain Adaptation in Predictive Healthcare via Concept-Grounded Orthogonal Inference
- FiSeR: Fine-Grained Source Representations for Cross-Domain AI Image Detection
and 16 more in the interactive map.
OOD and anomaly detection · 21 papers
- Less Precise Can Be More Reliable: A Systematic Evaluation of Quantization’s Impact on VLMs Beyond Accuracy (1 citation)
- A Geometry-Based View of Mahalanobis OOD Detection
- AdaSCALE: Adaptive Scaling for OOD Detection
- GEM-FI: Gated Evidential Mixtures with Fisher Modulation
- Geometrically Constrained Outlier Synthesis
- How Out-of-Distribution Detection Learning Theory Enhances Transformer: Learnability and Reliability
- Mechanistic Anomaly Detection via Functional Attribution
- Memory-Distilled Selection for Noise-Robust Anomaly Detection
- Mixture Prototype Flow Matching for Open-Set Supervised Anomaly Detection
- Towards One-for-All Anomaly Detection for Tabular Data
- A Close Look at Negative Label Guided Out-of-distribution Detection in Pre-trained Vision-Language Models
- Formally Exploring Visual Anomaly Detection Evaluation Metrics
and 9 more in the interactive map.
Semi-supervised learning · 21 papers
- Leveraging Lineage Barcodes as Natural Augmentations for Contrastive Learning of Cell Fate in scRNA-seq Data (1 citation)
- Are We Overconfident in Models and Results for Semi-Supervised 3D Medical Image Segmentation?
- Beyond Distribution Estimation: Simplex Anchored Structural Inference Towards Universal Semi-Supervised Learning
- Collaborative Learning for Semi-Supervised LiDAR Semantic Segmentation
- From Human Labels to Literature: Semi-Supervised Learning of NMR Chemical Shifts at Scale
- Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision
- Learning to Label: A Reinforced Self-Evolving Framework for Semi-supervised Referring Expression Segmentation
- Newton-coupled Dual-Teacher Semi-supervised Learning Framework
- Normality Calibration in Semi-supervised Graph Anomaly Detection
- Prediction-Powered Risk Monitoring of Deployed Models for Detecting Harmful Distribution Shifts
- Semi-LAR: Semi-supervised Contrastive Learning with Linear Attention for Removal of Nighttime Flares
- Semi-Supervised Gaze Estimation via Disentangled Subspace Contrastive Learning
and 9 more in the interactive map.
Noisy labels · 15 papers
- Data Augmentation of Contrastive Learning is Estimating Positive-incentive Noise (17 citations)
- Active Learning with Foundation Model Priors: Efficient Learning under Class Imbalance
- CARE: Class-Adaptive Expert Consensus for Reliable Learning with Long-Tailed Noisy Labels
- Intrinsic Gradient Suppression for Label-Noise Prompt Tuning in Vision–Language Models
- MIND: Decoupling Model-Induced Label Noise via Latent Manifold Disentanglement
- Stochastic Order Learning: An Approach to Rank Estimation Using Noisy Data
- The Label Horizon Paradox: Rethinking Supervision Targets in Financial Forecasting
- TPV: Parameter Perturbations Through the Lens of Test Prediction Variance
- GFD-EMVC: Evolutionary Multi-View Classification with Label Noise via Gradient and Feature Dual-Perception
- Just Y-Prediction: Enabling Historical Cumulative Inconsistency in Label Diffusion for Learning with Noisy Label
- Learning with Admissibility: Robust Fuzzy Hashing for Cross-Modal Retrieval with Noisy Labels
- MA$^3$S: Model-Agnostic Active Annotation Strategy for Crowdsourcing
and 3 more in the interactive map.
Test-time adaptation · 13 papers
- Coupled Training with Privileged Information and Unlabeled Data
- GFMate: Empowering Graph Foundation Models with Test-time Prompt Tuning
- Learning to Extrapolate to New Tasks: A Relational Approach to Task Extrapolation
- Mitigating Label Shift in Tabular In-Context Learning via Test-Time Posterior Adjustment
- Multi-Label Test-Time Adaptation with Bayesian Conditional Priors
- Tempora: Characterising the Time-Contingent Utility of Online Test-Time Adaptation
- USE: A Unified Self-Ensembling Framework for Test-Time Prompt Tuning
- Von Mises-Fisher Mixture Model with Dynamic Shrinkage for Realistic Test-Time Transduction
- Blocking the Leakage: Manifold-Aware Gradient Projection for Long-Horizon Test-Time Adaptation
- CONGA:Confidence-and-Gradient-Aware Learning Rate Schedule for Test Time Adaptation
- Learning to Memorize with Attributive and Associative Memory for Online Test-Time Adaptation of Vision-Language Models
- SaTeen: Learning Structural Alignment for Continual Test-Time Adaptation
and 1 more in the interactive map.
Knowledge distillation · 12 papers
- Generalist Graph Anomaly Detection via Prototype-Based Distillation
- Hard Labels In! Rethinking the Role of Hard Labels in Mitigating Local Semantic Drift
- Regret Pre-training: Bridging Prior and Posterior Views for Enhanced Knowledge Grounding
- The Bridge-Garden Dilemma in LLM Distillation: Why Mixing Hard and Soft Labels Works
- Toward Understanding Adversarial Distillation: Why Robust Teachers Fail
- Weak-to-Strong Generalization via Bregman Bias–Variance Decomposition
- Discrete Survival Knowledge Distillation for Competing Risks Analysis
- DomED: Redesigning Ensemble Distillation for Domain Generalization
- Don't Ignore the Tail: Decoupled Distillation Produces Top Maths Students on an Academic Budget
- Entropy-aware Span-Constrained Optimal Transport for Robust Cross-Tokenizer Knowledge Distillation
- Quantifying Cross-Domain Knowledge Distillation in the Presence of Domain Shift
- Why Self-Distillation Helps and Hurts: Denoising vs. Signal Forgetting