AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Test-Time Training with KV Binding Is Secretly Linear Attention

Junchen Liu, Sven Elflein, Or Litany, Zan Gojcic, Ruilong Li

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: State space and recurrent sequence models · 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
increase inner-loop iterations at inference (ablation)pretrained LaCT · LLM and NVSperplexity / PSNRHurt
replace gradient descent with gradient ascent in inner loop (retrained) (ablation)LaCT-LLM · Book-3 perplexityperplexity: 16.43 -> 16.19No clear effect
replace gradient descent with gradient ascent in inner loop (retrained) (ablation)ViTTT · ImageNet classificationtop-1 accuracy: 79.34 -> 79.61No clear effect
replace query with key at output (ablation)LaCT-NVS · NVSPSNR: 25.94 -> 25.95No clear effect
replace query with key at output (ablation)ViTTT · image classificationtop-1 accuracy: 79.34 -> 79.18No clear effect
update only last-layer inner-loop parameters (ablation)LaCT-LLM · Book-3 perplexityperplexity: 16.43 -> 15.93Improved
update only last-layer inner-loop parameters (ablation)ViTTT · image classificationtop-1 accuracy: 79.34% -> 79.63%Improved
remove weight normalization (ablation)LaCT-LLM Variant 1 · Book-3 perplexityperplexity: 15.93 -> 16.31Hurt
replace multi-layer MLP with single linear layer (ablation)LaCT-NVS · NVSPSNR: 25.93 -> 25.71Hurt
remove per-token learnable learning rate (ablation)LaCT-LLM Variant 3 · Book-3 perplexityperplexity: 16.23 -> 16.12No clear effect
remove momentum in inner-loop SGD (ablation)LaCT-LLM Variant 4 · Book-3 perplexityperplexity: 16.12 -> 15.97No clear effect
add gradient orthogonalization (ablation)linear attention (Variant 6) on LaCT-LLM · Book-3 perplexityperplexity: 16.80 -> 15.97Improved

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