Causally Sufficient and Necessary Feature Expansion for Class-Incremental Learning

2026-03-09

AI Summary

This work addresses feature collision and semantic confusion in class-incremental learning caused by spurious correlations both within and across tasks. From a causal perspective, it introduces the concept of Probability of Necessity and Sufficiency (PNS) into this domain for the first time and proposes a PNS-based regularization method. By designing a dual-network counterfactual generation mechanism, the approach jointly models the causal completeness of intra-task features and the separability of inter-task features to guide the feature expansion process. The proposed method significantly mitigates catastrophic forgetting, improves model accuracy and robustness across multiple benchmark datasets, and introduces the CPNS metric to quantitatively assess causal contributions.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Current expansion-based methods for Class Incremental Learning (CIL) effectively mitigate catastrophic forgetting by freezing old features. However, such task-specific features learned from the new task may collide with the old features. From a causal perspective, spurious feature correlations are the main cause of this collision, manifesting in two scopes: (i) guided by empirical risk minimization (ERM), intra-task spurious correlations cause task-specific features to rely on shortcut features. These non-robust features are vulnerable to interference, inevitably drifting into the feature space of other tasks; (ii) inter-task spurious correlations induce semantic confusion between visually similar classes across tasks. To address this, we propose a Probability of Necessity and Sufficiency (PNS)-based regularization method to guide feature expansion in CIL. Specifically, we first extend the definition of PNS to expansion-based CIL, termed CPNS, which quantifies both the causal completeness of intra-task representations and the separability of inter-task representations. We then introduce a dual-scope counterfactual generator based on twin networks to ensure the measurement of CPNS, which simultaneously generates: (i) intra-task counterfactual features to minimize intra-task PNS risk and ensure causal completeness of task-specific features, and (ii) inter-task interfering features to minimize inter-task PNS risk, ensuring the separability of inter-task representations. Theoretical analyses confirm its reliability. The regularization is a plug-and-play method for expansion-based CIL to mitigate feature collision. Extensive experiments demonstrate the effectiveness of the proposed method.

Problem

• Class-Incremental Learning
• Catastrophic Forgetting
• Feature Collision
• Spurious Correlations
• Causal Representation

Innovation

• Causal Inference
• Class-Incremental Learning
• Feature Expansion
• Probability of Necessity and Sufficiency
• Counterfactual Generation