AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Towards Foundation Models for Scientific Machine Learning: Characterizing Scaling and Transfer Behavior

Shashank Subramanian, Peter Harrington, Kurt Keutzer, Wahid Bhimji, Dmitriy Morozov, Michael W. Mahoney, Amir Gholami

NeurIPS 2023 · Architectures and efficiency › Pruning and efficient tuning · approach: Parameter-efficient transfer analysis · 14 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
pre-train on SYS-1(1,5) then fine-tune vs train from scratchFNO · SYS-1(5,10) downstreamexamples to reach error 1e-2: 8K -> 64Improved
pre-train and fine-tune vs scratch with large downstream data (~2^15 examples)FNO · SYS-1/SYS-2 downstreamtest errorNo clear effect
scale model from 64K to 256M parameters with fine-tuningFNO · SYS-1(5,10), SYS-2(1,2)test errorImproved
pre-train on a 64K-parameter modelFNO · SYS-1(5,10)test errorHurt
increase distribution shift of downstream physics to large OODFNO pre-trained · SYS-1(10,20)test errorMixed
pre-train on mixture of Poisson, Advection-Diffusion and HelmholtzFNO · SYS-1/2/3 downstreamtest errorImproved
remove per-instance input normalization (ablation)FNO · SYS-2performanceImproved

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