data attribution
The process of determining which parts of data contribute to a model's predictions, enhancing interpretability by identifying key features or inputs that drive decision-making.
- A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning
- A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning
- DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models
- Efficiently Verifiable Proofs of Data Attribution
- Fast Data Attribution for Text-to-Image Models
- Proxy-SPEX: Sample-Efficient Interpretability via Sparse Feature Interactions in LLMs
- Rescaled Influence Functions: Accurate Data Attribution in High Dimension
- Taming Hyperparameter Sensitivity in Data Attribution: Practical Selection Without Costly Retraining
- What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functions