SW-DRSO optimizes a tractable surrogate of worst-case expected loss over plausible inference-time corruptions using a barycentric adversary approximated via simplex weights.
The llama 3.1 series of models
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
representative citing papers
Embedding Arithmetic performs vector operations in the embedding space of T2I models to mitigate bias at inference time, outperforming baselines on diversity while preserving coherence via a new Concept Coherence Score.
BiasIG is a multi-dimensional benchmark for social biases in T2I models that shows debiasing interventions frequently cause confounding discrimination effects.
SynthPert fine-tunes LLMs using synthetic reasoning traces to reach state-of-the-art on the PerturbQA benchmark for cellular perturbation prediction, surpassing the generating frontier model while generalizing to unseen cell types with only 2% of filtered data.
A systematic review of T2I bias literature that distinguishes target and threshold fairness and proposes a target-based operationalization framework.
citing papers explorer
-
Distributionally Robust Set Representation Learning Under Inference-Time Element Corruption
SW-DRSO optimizes a tractable surrogate of worst-case expected loss over plausible inference-time corruptions using a barycentric adversary approximated via simplex weights.
-
Embedding Arithmetic: A Lightweight, Tuning-Free Framework for Post-hoc Bias Mitigation in Text-to-Image Models
Embedding Arithmetic performs vector operations in the embedding space of T2I models to mitigate bias at inference time, outperforming baselines on diversity while preserving coherence via a new Concept Coherence Score.
-
BiasIG: Benchmarking Multi-dimensional Social Biases in Text-to-Image Models
BiasIG is a multi-dimensional benchmark for social biases in T2I models that shows debiasing interventions frequently cause confounding discrimination effects.
-
SynthPert: Enhancing LLM Biological Reasoning via Synthetic Reasoning Traces for Cellular Perturbation Prediction
SynthPert fine-tunes LLMs using synthetic reasoning traces to reach state-of-the-art on the PerturbQA benchmark for cellular perturbation prediction, surpassing the generating frontier model while generalizing to unseen cell types with only 2% of filtered data.
-
Operationalizing Fairness in Text-to-Image Models: A Survey of Bias, Fairness Audits and Mitigation Strategies
A systematic review of T2I bias literature that distinguishes target and threshold fairness and proposes a target-based operationalization framework.