2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian Matching

[2506.05398] 2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian Matching Skip to main content Search Submit Donate Log in Search arXiv Press Enter to search · Advanced search Computer Science > Graphics arXiv:2506.05398 (cs) [Submitted on 3 Jun 2025 ( v1 ), last revised 28 Mar 2026 (this version, v2)] Title: 2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian Matching Authors: Caleb Zheng , Eli Shlizerman View a PDF of the paper titled 2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian Matching, by Caleb Zheng and 1 other authors View PDF HTML (experimental) Abstract: Diffusion models achieve remarkable performance across diverse generative tasks in computer vision, but their high computational cost remains a major barrier to deployment. Model pruning offers a promising way to reduce inference cost and enable lightweight models. However, pruning leads to quality drop due to reduced capacity. A key limitation of existing pruning approaches is that pruned models are finetuned using the same objective as the dense model (denoising score matching). Since the dense model is accessible during finetuning, it warrants a more effective approach for knowledge transfer from the dense to the pruned model. Motivated by this, we propose \textbf{2ndMatch} (\textbf{2ndM}), a general-purpose finetuning framework that introduces a \textbf{2nd}-order Jacobian ($J^{\top} J$) \textbf{M}atching loss inspired by Finite-Time Lyapunov Exponents. \textbf{2ndM} teaches the pruned model to mimic the sensitivity of the dense teacher, i.e., how to respond to small perturbations over time, through scalable random projections. The framework is architecture-agnostic and applies to both U-Net- and Transformer-based diffusion models. Experiments on CIFAR-10, CelebA, LSUN, ImageNet, and MSCOCO demonstrate that \textbf{2ndM} reduces the performance gap between pruned and dense models, substantially improving output quality. Comments: Accepted to CVPR 2026 Subjects: Graphics (cs.GR) Cite as: arXiv:2506.05398 [cs.GR] (or arXiv:2506.05398v2 [cs.GR] for this version) https://doi.org/10.48550/arXiv.2506.05398 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Caleb Zheng [ view email ] [v1] Tue, 3 Jun 2025 20:04:53 UTC (7,542 KB) [v2] Sat, 28 Mar 2026 00:58:33 UTC (13,282 KB) Full-text links: Access Paper: View a PDF of the paper titled 2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian Matching, by Caleb Zheng and 1 other authors View PDF HTML (experimental) TeX Source view license Current browse context: cs.GR < prev | next > new | recent | 2025-06 Change to browse by: cs References & Citations NASA ADS Google Scholar Semantic Scholar export BibTeX citation Loading... BibTeX formatted citation × loading... Data provided by: Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab) Major funding support from

Related documents

• Regularization Paths for Generalized Linear Models via Coordinate Descent.#6450
• Regularization Paths for Generalized Linear Models via Coordinate Descent#6408
• From Scores to Gibbs Correctors: Accelerating Uniform-Rate Discrete Diffusion Models#229480
• Particle GFlowNets: Rethinking Generative Marginalization Models#673530
• Why diffusion models do not memorize: the role of implicit dynamical regularization in training#1120376