Publications

• Language Models are Crossword Solvers

  Soumadeep Saha, S. Chakraborty, S. Saha, U. Garain
  NAACL (main) | 2025

• On Measuring Intrinsic Causal Attributions in Deep Neural Networks

  S. Saha, D.V. Rathore, Soumadeep Saha, D. Doermann, U. Garain
  CLeaR | 2025

• Analyzing Semantic Faithfulness of Language Models via Input Intervention on Question Answering

  A. Chaturvedi, S. Bhar, Soumadeep Saha, U. Garain, N. Asher
  Computational Linguistics | 2024

• VALUED - Vision and Logical Understanding Evaluation Dataset

  Soumadeep Saha, S. Saha, U. Garain
  DMLR (Data-centric Machine Learning Research) | 2024

• DOST—Domain Obedient Self-supervision for Trustworthy Multi-label Classification with Noisy Labels

  Soumadeep Saha, U. Garain, A. Ukil, A. Pal, S. Khandelwal
  W3PHIAI @ AAAI | 2024

• MedTric: A clinically applicable metric for evaluation of multi-label computational diagnostic systems

  Soumadeep Saha, U. Garain, A. Ukil, A. Pal, S. Khandelwal
  PLOS ONE | 2023

• LADDER: Revisiting the Cosmic Distance Ladder with Deep Learning Approaches and Exploring Its Applications

  R. Shah, Soumadeep Saha, P. Mukherjee, U. Garain, S. Pal
  The Astrophysical Journal Supplement Series (APJS) | 2024

Patents

• Method and system for contradiction avoided learning for multi-class multi-label classification

  A. Ukil, A. Pal, Soumadeep Saha, U. Garain
  US Patent | US12038949B2

• Method and system for evaluating clinical efficacy of multi-label multi-class computational diagnostic models

  A. Ukil, T. Deb, I. Sahu, S. C. Racha, S. Khandelwal, A. Pal, U. Garain, Soumadeep Saha
  US Patent (pending) | US20240096492A1