Pith. sign in

REVIEW 7 cited by

Dataset Distillation with Convexified Implicit Gradients

Not yet reviewed by Pith; the record is open.

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 2302.06755 v2 pith:OR6OQ5JN submitted 2023-02-13 cs.LG cs.CVstat.ML

Dataset Distillation with Convexified Implicit Gradients

classification cs.LG cs.CVstat.ML
keywords distillationdatasetgradientsimplicitalgorithmrcigstate-of-the-artconvexified
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We propose a new dataset distillation algorithm using reparameterization and convexification of implicit gradients (RCIG), that substantially improves the state-of-the-art. To this end, we first formulate dataset distillation as a bi-level optimization problem. Then, we show how implicit gradients can be effectively used to compute meta-gradient updates. We further equip the algorithm with a convexified approximation that corresponds to learning on top of a frozen finite-width neural tangent kernel. Finally, we improve bias in implicit gradients by parameterizing the neural network to enable analytical computation of final-layer parameters given the body parameters. RCIG establishes the new state-of-the-art on a diverse series of dataset distillation tasks. Notably, with one image per class, on resized ImageNet, RCIG sees on average a 108\% improvement over the previous state-of-the-art distillation algorithm. Similarly, we observed a 66\% gain over SOTA on Tiny-ImageNet and 37\% on CIFAR-100.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

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

  1. Robust Trajectory Distillation: Hybrid Reweighting Meets Teacher-Inspired Targets

    cs.CV 2026-06 unverdicted novelty 6.0

    Introduces SGR and TIAT for robust dataset distillation that suppresses noise while preserving knowledge under noisy supervision.

  2. Rank-Aware Hyperbolic Alignment for Vision-Language Dataset Distillation

    cs.CV 2026-06 unverdicted novelty 6.0

    RAHA applies rank-aware hyperbolic alignment to vision-language dataset distillation by enforcing geodesic alignment in the shared low-rank range and regularizing the residual subspace for improved transfer.

  3. Multimodal Distribution Matching for Vision-Language Dataset Distillation

    cs.CV 2026-05 unverdicted novelty 6.0

    MDM distills vision-language datasets via joint embedding clustering, weight-space model interpolation, and geometry-aware distribution matching on the unit hypersphere.

  4. DIVER:Diving Deeper into Distilled Data via Expressive Semantic Recovery

    cs.CV 2026-05 unverdicted novelty 6.0

    DIVER is a dual-stage distillation method using diffusion models to enhance semantic preservation and cross-architecture generalization in dataset distillation.

  5. DIVER:Diving Deeper into Distilled Data via Expressive Semantic Recovery

    cs.CV 2026-05 unverdicted novelty 5.0

    DIVER applies a pre-trained diffusion model in a dual-stage process of semantic inheritance, guidance, and fusion to improve semantic expression and cross-architecture generalization in dataset distillation.

  6. Federated Distillation for Whole Slide Image via Gaussian-Mixture Feature Alignment and Curriculum Integration

    cs.CV 2026-05 unverdicted novelty 5.0

    FedHD performs federated distillation for whole slide images by generating one synthetic feature set per real slide via Gaussian-mixture alignment and adding them via curriculum integration, outperforming prior federa...

  7. Federated Distillation for Whole Slide Image via Gaussian-Mixture Feature Alignment and Curriculum Integration

    cs.CV 2026-05 unverdicted novelty 5.0

    FedHD is a federated learning framework for whole slide images that distills one-to-one synthetic features aligned via Gaussian mixtures and progressively integrates cross-site features through curriculum learning to ...