4 papers
Semantic Density Effect (SDE): Maximizing Information Per Token Improves LLM Accuracy
Amr Ahmed
We introduce the Semantic Density Effect (SDE): the empirical finding that prompts carrying higher semantic information per token consistently produce more accurate, focused, and l…
Training-Only Heterogeneous Image-Patch-Text Graph Supervision for Advancing Few-Shot Learning Adapters
Mohammed Rahman Sherif Khan Mohammad, Ardhendu Behera, Sandip Pradhan +2
Recent adapter-based CLIP tuning (e.g., Tip-Adapter) is a strong few-shot learner, achieving efficiency by caching support features for fast prototype matching. However, these meth…
Inference-time Unlearning Using Conformal Prediction
Somnath Basu Roy Chowdhury, Rahul Kidambi, Avinava Dubey +4
Machine unlearning is the process of efficiently removing specific information from a trained machine learning model without retraining from scratch. Existing unlearning methods, w…
Fundamental Limits of Perfect Concept Erasure
Somnath Basu Roy Chowdhury, Avinava Dubey, Ahmad Beirami +4
Concept erasure is the task of erasing information about a concept (e.g., gender or race) from a representation set while retaining the maximum possible utility -- information from…