AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Enhancing Chemical Explainability Through Counterfactual Masking

Łukasz Janisiów, Marek Kochańczyk, Bartosz Michał Zieliński, Tomasz Danel

AAAI 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Machine-learned interatomic potentials · 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
replace feature zeroing with counterfactual masking (CReM) of non-shared fragmentsGIN solubility regressor · Common substructure pairs, single anchor|delta y| (lower better): 2.08 -> 1.68Improved
replace feature zeroing with CM (CReM)GIN solubility regressor · Common substructure pairs, multiple anchors|delta y| (lower better): 1.60 -> 1.32Improved
replace feature zeroing with CM (DiffLinker)GIN solubility regressor · Common substructure pairs, combined|delta y| (lower better): 1.80 -> 1.55Improved
use generative CM instead of feature zeroingmasking pipeline · Common substructure pairscoverage (validity): 100% -> 53% (multiple anchors, CReM)Hurt
mask with CM (CReM) instead of feature zeroingembeddings of masked molecules · Solubility test distributionWasserstein distance to test setImproved
use CM (CReM) instead of MMACE for counterfactual generationGIN classifiers on CYP3A4, CYP2D6, hERG · Counterfactual generationsynthetic accessibility (lower better): 3.69 -> 3.10 (hERG)Improved
use CM (CReM) instead of MMACEcounterfactual generation · CYP3A4, CYP2D6, hERGsimilarity to original: 0.65 -> 0.51 (hERG)Hurt
use DiffLinker instead of CReM as the generator (ablation)CM counterfactuals · CYP3A4, CYP2D6, hERGsynthetic accessibility: 2.71 -> 3.91 (CYP3A4)Hurt
use CM (CReM) instead of feature zeroing for benchmarking Grad-CAMXAI evaluation · Consistency score across 5 TDC datasetsconsistency score: 84 -> 71 (CYP2D6)Hurt
use CM (DiffLinker) for benchmarking explainersXAI evaluation · Consistency score across 5 datasetsconsistency score: 56 ± 28% (Grad-CAM CYP2D6) vs 50 ± 28% randomNo clear effect

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