Publications

Filter by type:
. Signal Propagation in Transformers: Theoretical Perspectives and the Role of Rank Collapse. NeurIPS 2022, 2022.

PDF

. Phenomenology of Double Descent in Finite-Width Neural Networks. ICLR 2022, 2021.

PDF

. Analytic Insights into Structure and Rank of Neural Network Hessian Maps . NeurIPS 2021, 2021.

PDF Code

. WoodFisher: Efficient second-order approximations for model compression. NeurIPS 2020, 2020.

PDF Code

. Model Fusion via Optimal Transport. NeurIPS 2020 (also, appeared at NeurIPS 2019, Optimal Transport & Machine Learning workshop), 2019.

PDF Code

. GLOSS: Generative Latent Optimization of Sentence Representations. ArXiv, 2019.

PDF

. Context Mover's Distance & Barycenters: Optimal transport of contexts for building representations.. AISTATS 2020 and ICLR 2019: DeepGenStruct workshop, 2019.

PDF

. Wasserstein is all you need. 2018.

Preprint

. RaaS and Hierarchical Aggregation Revisited. In ICWS, 2017.

PDF Code

. SL-FII: Syntactic and Lexical Constraints with Frequency based Iterative Improvement for Disease Mention Recognition in News Headlines. In IJCAI, 2016.

PDF Code