aiqbench analysis
This research introduces AIA², a novel framework designed to enhance AI model robustness against subgroup imbalances caused by data attributes like topics and demographics. By leveraging latent semantic distributions and large language models, the method improves performance on underrepresented subgroups without requiring explicit annotations. The study validates its effectiveness across diverse corpora, offering a practical solution for mitigating bias beyond traditional label imbalance issues.
AIA$^{2}$: Attribute-Agnostic Imbalance Augmentation for Subgroup Robustness
arXiv:2608.30297v1 Announce Type: new Abstract: Attributes describing data content and context can induce diverse imbalance patterns that go beyond label imbalance alone. However, existing studies primarily address label imbalance while overlooking data attributes, such as topics and demographics, which can induce meaningful subgroup structure while causing model degradation on underrepresented subgroups. We propose Attribute-Agnostic Imbalance Augmentation (AIA$^{2}$), a framework for improving model robustness under varying subgroup imbalances without explicit subgroup annotations. AIA$^{2}$ automatically discovers varying imbalances via latent semantic distributions, obtains slices with both learning difficulty and subgroup imbalance deficits, and deploys a large language model (LLM) for subgroup-aware imbalance augmentation. We have evaluated AIA$^{2}$ on 5 popular corpora with rich domains and their attribute values, covering social issues and diverse topics. Results show improved performance on the lowest-performing subgroups and consistent gains over competitive baselines. Ablation studies confirm complementary contributions from each component, and additional analyses show that AIA$^{2}$ provides a practical and consistent way to improve worst-group robustness under data subgroup imbalance. Code is available at https://github.com/trust-nlp/AIA2-Subgroup-Robustness.
Abstract:Attributes describing data content and context can induce diverse imbalance patterns that go beyond label imbalance alone. However, existing studies primarily address label imbalance while overlooking data attributes, such as topics and demographics, which can induce meaningful subgroup structure while causing model degradation on underrepresented subgroups. We propose Attribute-Agnostic Imbalance Augmentation (AIA$^{2}$), a framework for improving model robustness under varying subgroup imbalances without explicit subgroup annotations. AIA$^{2}$ automatically discovers varying imbalances via latent semantic distributions, obtains slices with both learning difficulty and subgroup imbalance deficits, and deploys a large language model (LLM) for subgroup-aware imbalance augmentation. We have evaluated AIA$^{2}$ on 5 popular corpora with rich domains and their attribute values, covering social issues and diverse topics. Results show improved performance on the lowest-performing subgroups and consistent gains over competitive baselines. Ablation studies confirm complementary contributions from each component, and additional analyses show that AIA$^{2}$ provides a practical and consistent way to improve worst-group robustness under data subgroup imbalance. Code is available at this https URL.
Computation and Language (cs.CL)
arXiv:2608.30297 [cs.CL]
(or arXiv:2608.30297v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2608.30297 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 261 words
Related coverage
Unified Semantic Modeling Framework for Large-Scale Job Understanding at LinkedIn
arXiv · Jul 29, 2026
CompDiff: Hierarchical Compositional Diffusion for Fair and Zero-Shot Intersectional Medical Image Generation
arXiv · Jul 20, 2026
Teach a Reward Model to Correct Itself: Reward Guided Adversarial Failure Discovery for Robust Reward Modeling
arXiv · Jun 8, 2026
SAGE: Subpopulation-Aware Generative Enhancement for Mitigating Spurious Correlations
arXiv · Sep 2, 2026