Usha Bhalla

14 papers A* 2Journal 11Unranked 1
YearRankTypeTitle / Venue / Authors
2026 — conf
Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders.
EACL (Volume 1: Long Papers)
Aaron J. Li, Suraj Srinivas, Usha Bhalla, Himabindu Lakkaraju
2025 J jnl
Building Bridges, Not Walls - Advancing Interpretability by Unifying Feature, Data, and Model Component Attribution.
CoRR
Shichang Zhang, Tessa Han, Usha Bhalla, Hima Lakkaraju
2025 J jnl
Interpretability Illusions with Sparse Autoencoders: Evaluating Robustness of Concept Representations.
CoRR
Aaron J. Li, Suraj Srinivas, Usha Bhalla, Himabindu Lakkaraju
2025 J jnl
RippleBench: Capturing Ripple Effects Using Existing Knowledge Repositories.
CoRR
Roy Rinberg, Usha Bhalla, Igor Shilov, Flávio P. Calmon, Rohit Gandikota
2025 J jnl
Temporal Sparse Autoencoders: Leveraging the Sequential Nature of Language for Interpretability.
CoRR
Usha Bhalla, Alex Oesterling, Claudio Mayrink Verdun, Himabindu Lakkaraju, Flávio P. Calmon
2024 J jnl
All Roads Lead to Rome? Exploring Representational Similarities Between Latent Spaces of Generative Image Models.
CoRR
Charumathi Badrinath, Usha Bhalla, Alex Oesterling, Suraj Srinivas, Himabindu Lakkaraju
2024 A* conf
Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE).
NeurIPS
Usha Bhalla, Alex Oesterling, Suraj Srinivas, Flávio P. Calmon, Himabindu Lakkaraju
2024 J jnl
Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE).
CoRR
Usha Bhalla, Alex Oesterling, Suraj Srinivas, Flávio P. Calmon, Himabindu Lakkaraju
2024 J jnl
Operationalizing the Blueprint for an AI Bill of Rights: Recommendations for Practitioners, Researchers, and Policy Makers.
CoRR
Alex Oesterling, Usha Bhalla, Suresh Venkatasubramanian, Himabindu Lakkaraju
2024 J jnl
Towards Unifying Interpretability and Control: Evaluation via Intervention.
CoRR
Usha Bhalla, Suraj Srinivas, Asma Ghandeharioun, Himabindu Lakkaraju
2023 A* conf
Discriminative Feature Attributions: Bridging Post Hoc Explainability and Inherent Interpretability.
NeurIPS
Usha Bhalla, Suraj Srinivas, Himabindu Lakkaraju
2023 J jnl
Do Vision-Language Pretrained Models Learn Composable Primitive Concepts?
Trans. Mach. Learn. Res.
Tian Yun, Usha Bhalla, Ellie Pavlick, Chen Sun
2023 J jnl
Verifiable Feature Attributions: A Bridge between Post Hoc Explainability and Inherent Interpretability.
CoRR
Usha Bhalla, Suraj Srinivas, Himabindu Lakkaraju
2022 J jnl
Do Vision-Language Pretrained Models Learn Primitive Concepts?
CoRR
Tian Yun, Usha Bhalla, Ellie Pavlick, Chen Sun