Virginia Smith

36 papers · 2014–2026 · 8 conferences · across top CS/AI conferences

Achievements

+11 more ↓

🐣 Hot Topic Early Bird 🌍 Conference Polyglot (6) 🧭 Keyword Pioneer 🌉 Interdisciplinary Bridge 🏃 Academic Marathon (11)

🧭 Keyword Pioneer 🐣 Hot Topic Early Bird 🐝 Cross-Pollinator (10) 🤝 Dynamic Duo (10) 👑 Triple Crown 🔥 Unstoppable (9) 💎 Century Club (34) ⚡ Prolific Year (5) 📈 Trend Setter ❓ The Questioner 🗃️ Keyword Collector (109)

Conferences

NIPS (13) ICLR (9) ICML (6) EMNLP (2) JMLR (2) NAACL (2) ACL (1) EACL (1)

Top co-authors

Tian Li (10) Amrith Setlur (7) Aashiq Muhamed (5) Mona T. Diab (5) Oscar Li (4) Shengyuan Hu (4) Maziar Sanjabi (4) Ahmad Beirami (4) Sergey Levine (3) Saurabh Garg (3)

Research topics

Differential Privacy (2) Privacy (2)

Keywords

federated learning (6) distribution shift (3) communication efficiency (3) differential privacy (3) distributed optimization (3) primal-dual optimization (2) few-shot learning (2) multi-task learning (2) retrieval-augmented generation (2) sparse autoencoder (2) large language model (2) machine learning (2) ensemble learning (2) text generation (2) transfer learning (1) data poisoning (1) contrastive learning (1) representation learning (1) semi-supervised learning (1) knowledge distillation (1)

Papers

Mechanistic Interpretability Should Prioritize Feature Consistency in Sparse Autoencoders ACL 2026 RefusalBench: Generative Evaluation of Selective Refusal in Grounded Language Models EACL 2026 CoRAG: Collaborative Retrieval-Augmented Generation NAACL 2025 Many-Objective Multi-Solution Transport ICLR 2025 Decoding Dark Matter: Specialized Sparse Autoencoders for Interpreting Rare Concepts in Foundation Models NAACL 2025 Semantic Agreement Enables Efficient Open-Ended LLM Cascades EMNLP 2025 Unlearning or Obfuscating? Jogging the Memory of Unlearned LLMs via Benign Relearning ICLR 2025 GRASS: Compute Efficient Low-Memory LLM Training with Structured Sparse Gradients EMNLP 2024 Prompting is a Double-Edged Sword: Improving Worst-Group Robustness of Foundation Models ICML 2024 On the Benefits of Public Representations for Private Transfer Learning under Distribution Shift NIPS 2024 RL on Incorrect Synthetic Data Scales the Efficiency of LLM Math Reasoning by Eight-Fold NIPS 2024 No Free Lunch in LLM Watermarking: Trade-offs in Watermarking Design Choices NIPS 2024 Complementary Benefits of Contrastive Learning and Self-Training Under Distribution Shift NIPS 2023 Bitrate-Constrained DRO: Beyond Worst Case Robustness To Unknown Group Shifts ICLR 2023 Progressive Ensemble Distillation: Building Ensembles for Efficient Inference NIPS 2023 Variance-Reduced Gradient Estimation via Noise-Reuse in Online Evolution Strategies NIPS 2023 On Tilted Losses in Machine Learning: Theory and Applications JMLR 2023 Differentially Private Adaptive Optimization with Delayed Preconditioners ICLR 2023 Diverse Client Selection for Federated Learning via Submodular Maximization ICLR 2022 On Privacy and Personalization in Cross-Silo Federated Learning NIPS 2022 Adversarial Unlearning: Reducing Confidence Along Adversarial Directions NIPS 2022 Label Leakage and Protection in Two-party Split Learning ICLR 2022 Private Adaptive Optimization with Side information ICML 2022 Tilted Empirical Risk Minimization ICLR 2021 Two Sides of Meta-Learning Evaluation: In vs. Out of Distribution NIPS 2021 Federated Hyperparameter Tuning: Challenges, Baselines, and Connections to Weight-Sharing NIPS 2021 Heterogeneity for the Win: One-Shot Federated Clustering ICML 2021 Ditto: Fair and Robust Federated Learning Through Personalization ICML 2021 On Large-Cohort Training for Federated Learning NIPS 2021 Fair Resource Allocation in Federated Learning ICLR 2020 Efficient Augmentation via Data Subsampling ICLR 2019 A Kernel Theory of Modern Data Augmentation ICML 2019 CoCoA: A General Framework for Communication-Efficient Distributed Optimization JMLR 2018 Federated Multi-Task Learning NIPS 2017 Adding vs. Averaging in Distributed Primal-Dual Optimization ICML 2015 Communication-Efficient Distributed Dual Coordinate Ascent NIPS 2014