Publications.

Reinforcement-learned unequal error protection for quantized semantic embeddings in bandwidth-constrained communication

FEHAN: A framework for explaining text classifiers through hierarchical attention networks

DICTA: Data-centric interpretable text analysis via surrogate models

HDExplain: High-dimensional explanations for graph neural network predictions

Task-driven control and model editing for deployed large language models

Towards interpretable molecular property prediction with subgraph-level explanations