AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Quantifying Cross-Attention Interaction in Transformers for Interpreting TCR-pMHC Binding

Jiarui Li, Zixiang Yin, Haley Smith, Zhengming Ding, Samuel J Landry, Ramgopal R. Mettu

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: Transformer neural operators for PDEs · 5 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
use QCAI with cross-attention instead of best self-attention baselineTULIP · TCR-XAI peptide perturbationLOdds: -0.736 -> -1.470Improved
use QCAI instead of RolloutTULIP · TCR-XAI CDR3a perturbationAOPC: 0.022 -> 0.014Hurt
use QCAI instead of best baselineTULIP · TCR-XAI BRHR top-25% CDR3ahit rate (%): 68.5 -> 79.1Improved
use QCAI instead of other methodsTULIP · TCR-XAI ROC-AUC peptideROC-AUC: 0.6024Improved
apply QCAI to cross-attention only vs both cross and self attention (ablation)QCAI · TCR-XAI CDR3b perturbationAOPC: 0.045 -> 0.048No clear effect
apply QCAI to self-attention only (Rollout) instead of cross-attention (ablation)QCAI · TCR-XAI peptide perturbationLOdds: -1.470 -> -0.044Hurt
use max instead of average aggregation across cross-attention layers (ablation)QCAI · TCR-XAI CDR3aBRHR.25 (%): 72.6 -> 79.1Improved
use max instead of average aggregation across cross-attention layers (ablation)QCAI · TCR-XAI peptideBRHR.25 (%): 76.7 -> 74.3Hurt
use QCAI instead of TokenTM on a larger integrated datasetTULIP · Integrated peptide perturbationAOPC: 0.026 -> 0.036Improved
replace faster baselines with QCAIQCAI · per-sample runtimems/sample: 0.11 -> 2.19Hurt

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