AI Paper Atlas

Research map / Large language models / AI and society

Who Gets Credit or Blame? Attributing Accountability in Modern AI Systems

Shichang Zhang, Hongzhe Du, Jiaqi W. Ma, Himabindu Lakkaraju

ICML 2026 · Large language models › AI and society · approach: Human-centered trust and agency · 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
use AA-Score instead of stage-level DVEmb for stage effect estimationstage effect estimator · MNIST MLP (average over 5 settings)correlation with retraining: 0.9417 -> 0.9456Improved
use AA-Score instead of DVEmbstage effect estimator · CelebA ResNet, Stage 3correlation with retraining: 0.4203 -> 0.8436Improved
use AA-Score instead of DVEmbstage effect estimator · CelebA ResNet, Stage 1correlation with retraining: 0.4406 -> 0.5524Improved
use AA-Score instead of DVEmbstage effect estimator · CivilComments Gemma-3, Stage 1correlation with retraining: 0.7766 -> 0.8576Improved
use AA-Score instead of DVEmbstage effect estimator · CivilComments Gemma-3, Stage 3correlation with retraining: 0.8514 -> 0.9746Improved
use AA-Score instead of DVEmbstage effect estimator · X-Ray DenseNet-121, Stage 1correlation with retraining: 0.9628 -> 0.9326Hurt
use AA-Score instead of DVEmbstage effect estimator · X-Ray DenseNet-121, Stage 3correlation with retraining: 0.7025 -> 0.7414Improved
use AA-Score instead of DVEmbstage effect estimator · MNIST MLP, Shift 2correlation with retraining: 0.9329 -> 0.9350Improved
use AA-Score instead of DVEmbstage effect estimator · MNIST MLP, Mislabelcorrelation with retraining: 0.9496 -> 0.9487No clear effect
skip the top two stages with positive effect on the confounderResNet on CelebA (blonde) · CelebA confounder log-likelihoodconfounder log-likelihood: decreases by 0.1130Improved
skip the bottom stage with negative effect on the confounderResNet on CelebA (blonde) · CelebA confounder log-likelihoodconfounder log-likelihood: increases by 0.0856Hurt
disperse poisoned examples evenly across all batches rather than concentrating themAA-Score · CIFAR-10 BadNets ResNet18poisoned batch detectionHurt

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