AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

Kernel Reboot: Breaking the Boundaries of Neural Tangent Kernels for Neural Fields

Amir Mallak, Alaa Maalouf, Lior Wolf, Daniela L. Rus, Dan Rosenbaum

TPAMI 2026 · Learning theory and algorithms › Kernels and regression theory · approach: Neural tangent kernel analysis · 1 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
add nonlinear refinement of delta-theta after KIP initialization (ablation)MetaQuill-KIP · MNIST held-out digitPSNR (dB): 11.31 -> 32.49Improved
initialize delta-theta from KIP support instead of KRR solution (ablation)MetaQuill-KIP · MNIST held-out digitPSNR (dB): 26.65 -> 32.49Improved
add nonlinear refinement beyond tangent-only meta adaptation (ablation)MetaQuill-KIP · MNIST held-out digitPSNR (dB): 18.32 -> 32.49Improved
add KIP-initialized nonlinear refinement over Single-INRMetaQuill-KIP · Flowers full reconstructionPSNR (dB): 13.57 -> 26.29Improved
add nonlinear refinement to KIP init (ablation)MetaQuill-KIP · Flowers full reconstructionPSNR (dB): 9.37 -> 26.29Improved
add nonlinear refinement to tangent-only adaptation (ablation)MetaQuill-KIP · Flowers full reconstructionPSNR (dB): 15.06 -> 26.29Improved
MetaQuill-KIP vs Single-INR(masked) at high observed fractionMetaQuill-KIP · Flowers inpainting (observed fraction 0.80)PSNR (dB): 14.00 -> 21.71Improved
MetaQuill-KIP vs Single-INR(masked) at very sparse observationMetaQuill-KIP · Flowers inpainting (observed fraction 0.10)PSNR (dB): 14.74 -> 11.97Hurt
MetaQuill-KIP vs Single-INR(masked) at 1% observedMetaQuill-KIP · Flowers inpainting (observed fraction 0.01)PSNR (dB): 13.60 -> 9.63Hurt
replace NTK regression with trained NF on circular-hole inpaintingNeural field · Flowers circular holePSNR (dB): 17.06 vs 20.21Mixed
use FPE instead of raw PE for NTK-KIP at high distillation (ablation)NTK-KIP · Flowers distillation 95%PSNR (dB): 19.81 -> 20.43Improved
use FPE instead of raw PE for NTK-KIP at 20% distillation (ablation)NTK-KIP · Flowers distillation 20%PSNR (dB): 33.35 -> 54.67Improved

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