Amit Sharma

Explaining machine learning classifiers through diverse counterfactual explanations

2020

Post-hoc explanations of machine learning models are crucial for people to understand and act on algorithmic predictions. An intriguing class of explanations is through counterfactuals, hypothetical examples that show people how to obtain a different …

Mental health in the global south

2019

Mental illness is rapidly gaining recognition as a serious global challenge. Recent human-computer interaction (HCI) research has investigated mental health as a domain of concern, but is yet to venture into the Global South, where …

Alleviating Privacy Attacks via Causal Learning

2019 · arXiv (Cornell University)

Machine learning models, especially deep neural networks have been shown to be susceptible to privacy attacks such as membership inference where an adversary can detect whether a data point was used for training a black-box …

Learnings from Technological Interventions in a Low Resource Language: A Case-Study on Gondi

2020 · arXiv (Cornell University)

The primary obstacle to developing technologies for low-resource languages is the lack of usable data. In this paper, we report the adoption and deployment of 4 technology-driven methods of data collection for Gondi, a low-resource …

Evaluating and Mitigating Bias in Image Classifiers: A Causal Perspective Using Counterfactuals

2022 · 2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)

Counterfactual examples for an input—perturbations that change specific features but not others—have been shown to be useful for evaluating bias of machine learning models, e.g., against specific demographic groups. However, generating counterfactual examples for images …

On Counterfactual Data Augmentation Under Confounding

2023 · arXiv (Cornell University)

Counterfactual data augmentation has recently emerged as a method to mitigate confounding biases in the training data. These biases, such as spurious correlations, arise due to various observed and unobserved confounding variables in the data …

DeduCE: Deductive Consistency as a Framework to Evaluate LLM Reasoning

2025 · arXiv (Cornell University)

Despite great performance on Olympiad-level reasoning problems, frontier large language models can still struggle on high school math when presented with novel problems outside standard benchmarks. Going beyond final accuracy, we propose a deductive consistency …