AIConfPaper
← Search

Suraj Srinivas

13 accepted papers

2026

Position: Explainability Research Must Prioritize Foundations over Ad-hoc Methods

Michal Moshkovitz, Suraj Srinivas, Lesia Semenova, Nave Frost, Cyrus Rashtchian, Valentyn Boreiko +4
ICML 2026poster

Despite the proliferation of Explainable AI (XAI) techniques—from feature attributions to sparse autoencoders—explanations rarely influence real-world workflows. In practice, they are often generated and discarded without guiding meaningful action. This gap reflects foundational shortcomings: resear…

Cited by 0Source ↗Scholar ↗
2025

How Much Can We Forget about Data Contamination?

Sebastian Bordt, Suraj Srinivas, Valentyn Boreiko, Ulrike von Luxburg
ICML 2025poster

The leakage of benchmark data into the training data has emerged as a significant challenge for evaluating the capabilities of large language models (LLMs). In this work, we challenge the common assumption that small-scale contamination renders benchmark evaluations invalid. First, we experimentally…

Cited by 1Source ↗PDF ↗code ↗Scholar ↗
2024

Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)

Usha Bhalla, Alex Oesterling, Suraj Srinivas, Flavio Calmon, Himabindu Lakkaraju
NeurIPS 2024poster

CLIP embeddings have demonstrated remarkable performance across a wide range of multimodal applications. However, these high-dimensional, dense vector representations are not easily interpretable, limiting our understanding of the rich structure of CLIP and its use in downstream applications that r…

Cited by 36Source ↗PDF ↗code ↗Scholar ↗
2023

Discriminative Feature Attributions: Bridging Post Hoc Explainability and Inherent Interpretability

Usha Bhalla, Suraj Srinivas, Himabindu Lakkaraju
NeurIPS 2023poster

With the increased deployment of machine learning models in various real-world applications, researchers and practitioners alike have emphasized the need for explanations of model behaviour. To this end, two broad strategies have been outlined in prior literature to explain models. Post hoc explanat…

Cited by 6Source ↗PDF ↗code ↗Scholar ↗
2023

On Minimizing the Impact of Dataset Shifts on Actionable Explanations

Anna P. Meyer, Dan Ley, Suraj Srinivas, Himabindu Lakkaraju
UAI 2023poster

The Right to Explanation is an important regulatory principle that allows individuals to request actionable explanations for algorithmic decisions. However, several technical challenges arise when providing such actionable explanations in practice. For instance, models are periodically retrained to…

Cited by 6Source ↗PDF ↗Scholar ↗
2023

Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustness

Suraj Srinivas, Sebastian Bordt, Himabindu Lakkaraju
NeurIPS 2023spotlight

One of the remarkable properties of robust computer vision models is that their input-gradients are often aligned with human perception, referred to in the literature as perceptually-aligned gradients (PAGs). Despite only being trained for classification, PAGs cause robust models to have rudimentary…

Cited by 13Source ↗PDF ↗code ↗Scholar ↗
2022

Data-Efficient Structured Pruning via Submodular Optimization

Marwa El Halabi, Suraj Srinivas, Simon Lacoste-Julien
NeurIPS 2022accept

Structured pruning is an effective approach for compressing large pre-trained neural networks without significantly affecting their performance. However, most current structured pruning methods do not provide any performance guarantees, and often require fine-tuning, which makes them inapplicable in…

Cited by 22Source ↗PDF ↗code ↗Scholar ↗
2022

Efficient Training of Low-Curvature Neural Networks

Suraj Srinivas, Kyle Matoba, Himabindu Lakkaraju, François Fleuret
NeurIPS 2022accept

Standard deep neural networks often have excess non-linearity, making them susceptible to issues such as low adversarial robustness and gradient instability. Common methods to address these downstream issues, such as adversarial training, are expensive and often sacrifice predictive accuracy. In…

Cited by 25Source ↗PDF ↗Scholar ↗
2022

Which Explanation Should I Choose? A Function Approximation Perspective to Characterizing Post Hoc Explanations

Tessa Han, Suraj Srinivas, Himabindu Lakkaraju
NeurIPS 2022accept

A critical problem in the field of post hoc explainability is the lack of a common foundational goal among methods. For example, some methods are motivated by function approximation, some by game theoretic notions, and some by obtaining clean visualizations. This fragmentation of goals causes not on…

Cited by 104Source ↗PDF ↗code ↗Scholar ↗
2021

Rethinking the Role of Gradient-based Attribution Methods for Model Interpretability

Suraj Srinivas, Francois Fleuret
ICLR 2021oral

Current methods for the interpretability of discriminative deep neural networks commonly rely on the model's input-gradients, i.e., the gradients of the output logits w.r.t. the inputs. The common assumption is that these input-gradients contain information regarding $p_{\theta} ( y\mid \mathbf{x} )…

Cited by 56Source ↗PDF ↗code ↗Scholar ↗