Increase trust in AI systems

Just like medical treatments or economic policies, AI systems can be considered as interventions in our society. I build methods to help experts interpret an AI system’s output and estimate their (differential) impact on different subpopulations. One of these methods, Diverse Counterfactual Explanations (DiCE), has been integrated as a part of Microsoft’s Responsible AI platform.

I also work on improving the generalizability of ML models so that they are less sensitive to their training data distribution. This line of work has led to a NASSCOM AI GameChangers Award (2023-24).

Counterfactual explanations
Building prediction models that generalize better
  • The challenges of prediction and explanation in social systems (Hofman et al., 2017)
  • Using causal reasoning to build generalizable ML classifiers (missing reference)
Role of causality in trustworthy ML
  • The necessary role of causality in understanding when ML explanations can improve human understanding (Chen et al., 2023)
  • A framework for deploying trustworthy ML systems based on technology readiness levels (Lavin et al., 2022)
  • ML fairness estimates can be misleading without modeling data missingness (Goel et al., 2021)

References

  1. FAccT 2020
    Explaining machine learning classifiers through diverse counterfactual explanations
    Ramaravind K Mothilal, Amit Sharma, and Chenhao Tan
    Proceedings of the 2020 ACM conference on Fairness, Accountability and Transparency (FAccT), Mar 2020
  2. ICLR 2024
    Faithful Explanations of Black-box NLP Models Using LLM-generated Counterfactuals
    Yair Ori Gat, Nitay Calderon, Amir Feder, Alexander Chapanin, Amit Sharma, and Roi Reichart
    In The Twelfth International Conference on Learning Representations (ICLR), Mar 2024
  3. AIES 2021
    Towards Unifying Feature Attribution and Counterfactual Explanations: Different Means to the Same End
    Ramaravind Kommiya Mothilal, Divyat Mahajan, Chenhao Tan, and Amit Sharma
    In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, Jul 2021
  4. WACV 2020
    Evaluating and mitigating bias in image classifiers: A causal perspective using counterfactuals
    Saloni Dash, V Balasubramanian, and Amit Sharma
    Proc. IEEE Workshop Appl. Comput. Vis., Sep 2020
  5. Science
    Prediction and explanation in social systems
    Jake M Hofman, Amit Sharma, and Duncan J Watts
    In Science, Sep 2017
  6. TMLR 2023
    Machine Explanations and Human Understanding
    Chacha Chen, Shi Feng, Amit Sharma, and Chenhao Tan
    Transactions on Machine Learning Research, Sep 2023
  7. Nature Commun.
    Technology readiness levels for machine learning systems
    Alexander Lavin, Ciarán M Gilligan-Lee, Alessya Visnjic, Siddha Ganju, Dava Newman, Sujoy Ganguly, Danny Lange, Atílím Güneş Baydin, Amit Sharma, Adam Gibson, Stephan Zheng, Eric P Xing, Chris Mattmann, James Parr, and Yarin Gal
    Nature Communications, Oct 2022
  8. AAAI 2021
    The importance of modeling data missingness in algorithmic fairness: A causal perspective
    N Goel, A Amayuelas, A Deshpande, and  others
    Proceedings of the AAAI Conference on Artificial Intelligence, Oct 2021