A Framework for Task-Agnostic Counterfactual Fairness in Transformers through Activation Steering
cs.LG, cs.AI, cs.CY
Submitted: 2026-10-30
Updated: 2026-10-30
Terminology
Sources
- Mechanistic Interpretability for AI Safety -- A Review
- A Survey on Large Language Models for Critical Societal Domains: Finance, Healthcare, and Law
- Efficient and Interpretable Transformer for Counterfactual Fairness
- Toy Models of Superposition
- How to use and interpret activation patching
- The Dangers of Post-hoc Interpretability: Unjustified Counterfactual Explanations
- The Linear Representation Hypothesis and the Geometry of Large Language Models
- Identifiability of Causal Graphs using Functional Models
- FairPFN: Transformers Can do Counterfactual Fairness
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis