Shapley Values
PulseAugur coverage of Shapley Values — every cluster mentioning Shapley Values across labs, papers, and developer communities, ranked by signal.
-
New method combats false-name manipulation in ML data attribution
Researchers have developed a new data attribution method called the quotient semivalue mechanism to combat false-name manipulation in machine learning. This approach addresses issues where contributors might inflate the…
-
QuadraSHAP offers stable, scalable Shapley values for product games
Researchers have developed QuadraSHAP, a novel method for efficiently calculating Shapley values in product games, which are common in machine learning explainability. The technique reduces the complex calculation to a …
-
Quadrature-TreeSHAP 提供更快、更稳定的AI模型解释
研究人员开发了Quadrature-TreeSHAP，一种用于解释树集成预测的新颖方法，该方法独立于深度，并且比现有方法更具数值稳定性。这项新技术自然地扩展到高阶Shapley交互值，并利用基于求积的重构来实现高效计算。实证评估表明，Quadrature-TreeSHAP在Shapley值和交互计算的速度方面均显著优于TreeSHAP和GPUTreeSHAP。
-
Researchers develop stable, explainable AI for elderly fall detection
Researchers have developed a new framework for skeleton-based fall detection that uses a temporally stabilized attribution mechanism called T-SHAP. This method enhances the interpretability of AI models used in elderly …
-
New statistical viewpoint improves probabilistic value estimation for AI
Researchers have developed a new statistical viewpoint for understanding and improving probabilistic value estimation methods. Their work identifies a common first-order error structure across existing estimators, which…
-
New framework uses K-Shapley values for meritocratic fairness in bandits
Researchers have introduced a novel framework for achieving meritocratic fairness in budgeted combinatorial multi-armed bandits with full-bandit feedback. This new approach extends the Shapley value concept to a K-Shapl…
-
Researchers develop Shapley value explainers for temporal graph neural networks
Researchers have developed two new model-agnostic explainers for Temporal Graph Neural Networks (TGNNs), utilizing Shapley and Owen values. These methods aim to make the predictions of TGNNs, which combine spatial and t…
-
Study finds Shapley value benchmarks for AI explainability misaligned with human utility
A new paper examines the evaluation of explainable AI (XAI) methods, specifically Shapley value variants, in high-stakes scenarios like fraud detection. Researchers found that standard quantitative metrics for XAI do no…