Research map / Label-efficient and robust learning / Self-supervised and contrastive
Predicting Visual Political Bias Using Webly Supervised Data and an Auxiliary Task
Christopher Thomas, Adriana Kovashka
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 |
|---|---|---|---|
| train with latent text document embedding as privileged information | ResNet political bias classifier · topic Vaccines | average performance (accuracy): 0.6768 -> 0.7422 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · topic Climate Change | average performance (accuracy): 0.6934 -> 0.7179 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · top-20 media sources | average performance (accuracy): 0.697 -> 0.739 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · top-100 media sources | average performance (accuracy): 0.690 -> 0.724 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · Sun | average performance (accuracy): 0.627 -> 0.707 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · Breitbart | average performance (accuracy): 0.527 -> 0.607 | Improved |
| train with latent text document embedding as privileged information | ResNet political bias classifier · RevCom | average performance (accuracy): 0.875 -> 0.781 | Hurt |
| train with latent text document embedding as privileged information | ResNet political bias classifier · Slate | average performance (accuracy): 0.798 -> 0.793 | No clear effect |
Closest papers by meaning
- Classifier-to-Bias: Toward Unsupervised Automatic Bias Detection for Visual Classifiers (CVPR 2025 · similarity 0.82)
- Understanding Bias in Large-Scale Visual Datasets (NeurIPS 2024 · similarity 0.81)
- Political Bias Prediction Models Focus on Source Cues, Not Semantics (AAAI 2025 · similarity 0.79)
- Discovering and Mitigating Visual Biases through Keyword Explanation (CVPR 2024 · similarity 0.79)
- ConceptScope: Characterizing Dataset Bias via Disentangled Visual Concepts (NeurIPS 2025 · similarity 0.79)
- Quantifying Political Polarization through the Lens of Machine Translation and Vicarious Offense (AAAI 2024 · similarity 0.79)
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.
