AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Generative Adaptation of Dynamics to Environmental Shifts via Weight-space Diffusion

Ruikun Li, Huandong Wang, Jingtao Ding, Yuan Yuan, Qingmin Liao, Yong Li

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Low-rank compression and merging · 0 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
generate full expert weights with weight-space diffusion (DynaDiff) instead of meta-learning modulationDynaDiff · Cylinder Flow out-domainRMSE: 0.082 (GEPS) -> 0.065Improved
use 1M-parameter generated experts instead of ~500M foundation modelsDynaDiff · Navier-Stokes out-domainRMSE: 0.074 (FNO) -> 0.064Improved
use DynaDiff weight graph, functional loss and prompter instead of D2NWG latent weight diffusionweight-space generation · Lambda-Omega out-domainRMSE: 0.105 -> 0.091Improved
generate weights with CVAE instead of diffusionweight-space generation · Kolmogorov FlowRMSE: 4.454 in-domain, 29.089 out-domainHurt
add function loss to VAE training (ablation)DynaDiff · Kolmogorov Flow and Navier-Stokesprediction accuracyImproved
add domain-adaptive initialization when building the model zoo (ablation)DynaDiff · Kolmogorov Flow and Navier-Stokesprediction accuracyImproved
reduce model zoo size below 50 (ablation)DynaDiff · Cylinder Flow and Lambda-Omegaprediction errorHurt
increase number of seen environments beyond about 20% of total (ablation)DynaDiff · Kolmogorov Flow and Navier-Stokesprediction errorNo clear effect
randomly permute 10% of neurons in the model zoo (ablation)DynaDiff weight graph · Cylinder FlowperformanceNo clear effect
swap FNO expert for WNO or UNO (ablation)DynaDiff · Cylinder Flowprediction performanceNo clear effect
train prompter with variable observation length 1 to LDynaDiff prompter · observation length sensitivityprediction qualityImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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