Research map / Probabilistic and scientific modeling / Molecules and proteins
Priority-Aware Shapley Value
Kiljae Lee, Ziqi Liu, Weijing Tang, Yuan Zhang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use a general DAG with extra domain relations instead of a coarse ordered partition | PSV (PASV with uniform weights) · Census Income feature attribution | feature attribution values | Mixed |
| increase base b of priority weights for Anchor/Copier/Poisoner | PASV · MNIST data valuation (8 providers) | provider-level value | Improved |
| replace classical SV with precedence-aware PSV | Shapley valuation · MNIST data valuation (8 providers) | credit to copied data | Improved |
| approximate the DAG by a two-layer ordered partition (WSV) instead of the full DAG | PASV · MNIST data valuation (8 providers) | provider-level value | Hurt |
| add quadratic regression proxy with bias correction to the direct MC estimator | PASV Monte Carlo estimator · Synthetic sum-of-unanimity utility | approximation error: n <= 512 | Mixed |
| make priority weights lambda non-uniform (Unif(1,R), R up to 1024) | adjacent-swap M-H sampler · Synthetic poset families | empirical mixing time: R up to 1024 | No clear effect |
| perturb single DAG edges (deletion or addition) (ablation) | PASV on Census Income · Census Income DAG sensitivity | RAS: RAS below 0.05; Pearson above 0.97 | No clear effect |
Closest papers by meaning
- Generalized Priority-Aware Shapley Value (NeurIPS 2026 · similarity 0.88)
- DU-Shapley: A Shapley Value Proxy for Efficient Dataset Valuation (NeurIPS 2024 · similarity 0.87)
- Incorporating Information into Shapley Values: Reweighting via a Maximum Entropy Approach (ICML 2024 · similarity 0.87)
- A Unified Framework for Provably Efficient Algorithms to Estimate Shapley Values (NeurIPS 2025 · similarity 0.87)
- Rethinking Data Shapley for Data Selection Tasks: Misleads and Merits (ICML 2024 · similarity 0.86)
- Regression-adjusted Monte Carlo Estimators for Shapley Values and Probabilistic Values (NeurIPS 2025 · similarity 0.86)
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.
