Pith. sign in

REVIEW 26 cited by

Contrastive Learning with Hard Negative Samples

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2010.04592 v2 pith:SVDFBL5Y submitted 2020-10-09 cs.LG stat.ML

Joshua Robinson , Ching-Yao Chuang , Suvrit Sra , Stefanie Jegelka This is my paper
classification cs.LGstat.ML
keywords negativecontrastivehardlearningsamplessamplingmethodsunsupervised
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

How can you sample good negative examples for contrastive learning? We argue that, as with metric learning, contrastive learning of representations benefits from hard negative samples (i.e., points that are difficult to distinguish from an anchor point). The key challenge toward using hard negatives is that contrastive methods must remain unsupervised, making it infeasible to adopt existing negative sampling strategies that use true similarity information. In response, we develop a new family of unsupervised sampling methods for selecting hard negative samples where the user can control the hardness. A limiting case of this sampling results in a representation that tightly clusters each class, and pushes different classes as far apart as possible. The proposed method improves downstream performance across multiple modalities, requires only few additional lines of code to implement, and introduces no computational overhead.

Discussion (0). Sign in to comment.

Forward citations

Cited by 26 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Fusion Embedding: A Unified Embedding Space for Text, Image, Video, and Audio

    cs.CL 2026-07 conditional novelty 7.0 of 10

    Trained connectors and audio-only gated adapters integrate audio into a frozen vision-language embedding space, preserving base outputs bit-exactly and yielding emergent audio-image retrieval.

  2. Reasoning Text-to-Video Retrieval for Operating Room Clips via Action-Driven Digital Twins

    cs.CV 2026-06 conditional novelty 7.0 of 10

    OR3 converts OR clips to action-driven digital twins, uses LLM imagination for hypothetical ActDTs, and achieves 57.6 R@1 and 77.3 R@5 on 276 implicit queries from 386 robotic knee procedure clips, outperforming baselines.

  3. Chameleon: Style-Content Disentangled Framework for Cross-Domain Object Compositing

    cs.CV 2026-05 unverdicted novelty 7.0 of 10

    Chameleon proposes the first large-scale cross-domain compositing dataset and a disentangled encoder plus gated diffusion transformer that outperforms prior in-domain and cross-domain methods on plausibility and fidelity.

  4. Contrast to Detect: Dynamic Graph Contrastive Regularization for Unsupervised Anomaly Detection in Multivariate Time Series

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    ContrastAD achieves highest mean F1 on all five MTS benchmarks and highest AUC on three by building DTW-based sparse graph snapshots and contrasting divergent pairs with a stable anchor instead of enforcing invariance.

  5. Learning over Positive and Negative Edges with Contrastive Message Passing

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    Contrastive Message Passing lets GNNs apply similarity-preserving transforms to positive edges and dissimilarity-inducing transforms to negative edges via soft positive semidefinite constraints on weights, yielding ga...

  6. MASS-DPO: Multi-negative Active Sample Selection for Direct Policy Optimization

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    MASS-DPO derives a Plackett-Luce-specific log-determinant Fisher information objective to select non-redundant negative samples, matching or exceeding multi-negative DPO performance with substantially fewer negatives ...

  7. DiffusionPrint: Learning Generative Fingerprints for Diffusion-Based Inpainting Localization

    cs.CV 2026-04 unverdicted novelty 7.0 of 10

    DiffusionPrint learns robust forensic feature maps via MoCo-style contrastive training on diffusion inpainting fingerprints, boosting localization accuracy by up to 28% when fused into existing IFL systems and general...

  8. Lighting-Consistent Object Transfer Across Radiance Fields

    cs.GR 2026-06 unverdicted novelty 6.0 of 10

    Diffusion-based per-view harmonization for lighting-consistent object transfer between 3DGS scenes, using heterogeneous training data and final 3D consolidation.

  9. Doing well with less! On Sampling Techniques for Empirical Pairwise Loss Estimation/Minimization

    stat.ML 2026-06 unverdicted novelty 6.0 of 10

    Sampling pairs directly with auxiliary information for higher inclusion probabilities on informative pairs yields near-full pairwise loss performance at reduced computational cost.

  10. Beyond Topical Similarity: Contrastive Evidence Retrieval with Interpretable Attention Alignment in RAG

    cs.CL 2026-05 unverdicted novelty 6.0 of 10

    CERA fine-tunes a dense retriever with triplet contrastive learning plus attention alignment to human rationales, claiming better retrieval effectiveness and faithfulness on clinical trial reports than Contriever and ...

  11. HOLA: Holistic Multi-Modal Alignment for Open-Set 3D Recognition

    cs.CV 2026-05 unverdicted novelty 6.0 of 10

    HOLA introduces multi-view multi-text alignment and a decoupled contrastive loss for state-of-the-art open-vocabulary 3D recognition on long-tail benchmarks.

  12. Generalizable Object Re-Identification via Visual In-Context Prompting

    cs.CV 2025-08 conditional novelty 6.0 of 10

    VICP uses an LLM to generate per-category visual prompts for a frozen DINOv2, enabling few-shot generalization to unseen object categories in re-identification without parameter updates.

  13. QuRe: Query-Relevant Retrieval through Hard Negative Sampling in Composed Image Retrieval

    cs.CV 2025-07 conditional novelty 6.0 of 10

    QuRe trains composed image retrieval models with a pairwise reward objective on hard negatives found between sharp relevance-score drops, and adds a human-preference benchmark for evaluating retrieval relevance.

  14. From Exploration to Revelation: Detecting Dark Patterns in Mobile Apps

    cs.SE 2024-11 unverdicted novelty 6.0 of 10

    AppRay integrates LLM-guided task-oriented exploration with a contrastive learning multi-label classifier and rule-based refiner to detect intra- and inter-page dark patterns, reporting 0.89/0.85 F1 on new datasets wi...

  15. Multimodal Semantic-Aware Contrastive Learning For False Negative Mitigation in 3D Medical Imaging

    cs.LG 2026-07 conditional novelty 5.0 of 10

    MseaCL, a semantic-aware contrastive pretraining method for 3D brain MRI and reports, reports a 0.226 external AUC gain over instance-based contrastive learning for pediatric brain tumor molecular classification.

  16. Aligning Implied Statements for Implicit Hate Speech Generalizability with Context-Bounded Semi-hard Negative Mining

    cs.CL 2026-06 unverdicted novelty 5.0 of 10

    ImpSH improves cross-domain generalization in implicit hate speech classification by aligning posts with implied statements and applying context-bounded semi-hard negative mining within a triplet learning setup.

  17. SurfSurg6D: Geometry Consistent Dense Correspondence for Textureless Surgical Instrument Pose Estimation

    cs.CV 2026-05 unverdicted novelty 5.0 of 10

    A new synthetic dataset and geometry-consistent dense correspondence framework improve RGB-only pose estimation accuracy for surgical instruments on three evaluation datasets.

  18. MSAlign: Aligning Molecule and Mass Spectra Foundation Models for Metabolite Identification

    cs.LG 2026-05 conditional novelty 5.0 of 10

    MSAlign aligns frozen DreaMS and ChemBERTa models with MLPs and candidate-based contrastive learning to outperform prior methods on molecule retrieval from MS/MS spectra while quantifying distribution shift in data splits.

  19. Combating Visual Neglect and Semantic Drift in Large Multimodal Models for Enhanced Cross-Modal Retrieval

    cs.CV 2026-04 unverdicted novelty 5.0 of 10

    SSA-ME uses saliency-aware modeling to reduce visual neglect and semantic drift, achieving SOTA results on the MMEB benchmark for multimodal retrieval.

  20. Concrete Jungle: Towards Concreteness Paved Contrastive Negative Mining for Compositional Understanding

    cs.LG 2026-04 unverdicted novelty 5.0 of 10

    Using lexical concreteness to guide contrastive negative mining and a new margin-based Cement loss, the Slipform framework reaches state-of-the-art on compositional benchmarks for vision-language models.

  21. A Generalized Learning Framework for Self-Supervised Contrastive Learning

    cs.LG 2025-08 unverdicted novelty 5.0 of 10

    A single framework unifies BYOL, Barlow Twins, and SwAV, plus a plug-in calibration method, ADC, that improves learned representations by preserving input-space distances.

  22. Parameter-Efficient Adaptation of SAM 3 for Automated ITV Generation from 4DCT Images

    eess.IV 2026-06 unverdicted novelty 4.0 of 10

    LoRA-adapted SAM 3 with hard-negative mining and phase-coherent filtering achieves median Dice 0.968 on pulmonary structures from 4DCT using seven annotated volumes.

  23. Similarity Choice and Negative Scaling in Supervised Contrastive Learning for Deepfake Audio Detection

    eess.AS 2026-04 unverdicted novelty 4.0 of 10

    Cosine similarity in SupCon with a delayed negative queue on wav2vec2 XLS-R yields the lowest equal error rates for deepfake audio detection on in-the-wild and pooled evaluations.

  24. The Impact of Semantic Pairs on Self-Supervised Representation Learning

    cs.LG 2025-10 conditional novelty 4.0 of 10

    Semantic positive pairs (same-class, different images) consistently beat augmented-view pairs for self-supervised pretraining in controlled ImageNet-subset experiments, with the largest gains for SimCLR.

  25. eMargin: Revisiting Contrastive Learning with Margin-Based Separation

    cs.LG 2025-07 reject novelty 4.0 of 10

    An adaptive margin added to InfoNCE improves time series clustering metrics but hurts linear-probe classification, exposing a disconnect between clustering scores and downstream utility.

  26. Subject Invariant Contrastive Learning for Human Activity Recognition

    cs.CV 2025-07 conditional novelty 4.0 of 10

    A subject-reweighted contrastive loss improves cross-subject generalization for human activity recognition across unimodal, multimodal, and supervised settings.

Pith tools