Ashwinee Panda

50 papers A* 9A 1Journal 38Unranked 2

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	@GrokSet: multi-party Human-LLM Interactions in Social Media. CoRR Matteo Migliarini, Berat Ercevik, Oluwagbemike Olowe, Saira Fatima, Sarah Zhao, Minh Anh Le, Vasu Sharma, Ashwinee Panda
2026	J	`jnl`	A Few Bad Neurons: Isolating and Surgically Correcting Sycophancy. CoRR Claire O'Brien, Jessica Seto, Dristi Roy, Aditya Dwivedi, Sunishchal Dev, Kevin Zhu, Sean O'Brien, Ashwinee Panda, Ryan Lagasse
2026	J	`jnl`	From Refusal Tokens to Refusal Control: Discovering and Steering Category-Specific Refusal Directions. CoRR Rishab Alagharu, Ishneet Sukhvinder Singh, Shaibi Shamsudeen, Zhen Wu, Ashwinee Panda
2026	J	`jnl`	Multi-Token Prediction via Self-Distillation. CoRR John Kirchenbauer, Abhimanyu Hans, Brian R. Bartoldson, Micah Goldblum, Ashwinee Panda, Tom Goldstein
2026	J	`jnl`	Speculating Experts Accelerates Inference for Mixture-of-Experts. CoRR Vivan Madan, Prajwal Singhania, Abhinav Bhatele, Tom Goldstein, Ashwinee Panda
2025	J	`jnl`	Alignment-Constrained Dynamic Pruning for LLMs: Identifying and Preserving Alignment-Critical Circuits. CoRR Dev Patel, Gabrielle Gervacio, Diekola Raimi, Kevin Zhu, Ryan Lagasse, Gabriel Grand, Ashwinee Panda, Maheep Chaudhary
2025	J	`jnl`	Amortized Latent Steering: Low-Cost Alternative to Test-Time Optimization. CoRR Nathan Egbuna, Saatvik Gaur, Sunishchal Dev, Ashwinee Panda, Maheep Chaudhary
2025	J	`jnl`	Analysis of Attention in Video Diffusion Transformers. CoRR Yuxin Wen, Jim Wu, Ajay Jain, Tom Goldstein, Ashwinee Panda
2025	J	`jnl`	Continual Pre-training of MoEs: How robust is your router? CoRR Benjamin Thérien, Charles-Étienne Joseph, Zain Sarwar, Ashwinee Panda, Anirban Das, Shi-Xiong Zhang, Stephen Rawls, Sambit Sahu, Eugene Belilovsky, Irina Rish
2025	J	`jnl`	Continual Pre-training of MoEs: How robust is your router? Trans. Mach. Learn. Res. Benjamin Thérien, Charles-Étienne Joseph, Zain Sarwar, Ashwinee Panda, Anirban Das, Shi-Xiong Zhang, Stephen Rawls, Sambit Sahu, Eugene Belilovsky, Irina Rish
2025	J	`jnl`	Dense Backpropagation Improves Training for Sparse Mixture-of-Experts. CoRR Ashwinee Panda, Vatsal Baherwani, Zain Sarwar, Benjamin Thérien, Supriyo Chakraborty, Tom Goldstein
2025	J	`jnl`	Differentially Private Image Classification by Learning Priors from Random Processes. J. Priv. Confidentiality Xinyu Tang, Ashwinee Panda, Vikash Sehwag, Prateek Mittal
2025	J	`jnl`	DynaGuard: A Dynamic Guardrail Model With User-Defined Policies. CoRR Monte Hoover, Vatsal Baherwani, Neel Jain, Khalid Saifullah, Joseph Vincent, Chirag Jain, Melissa Kazemi Rad, C. Bayan Bruss, Ashwinee Panda, Tom Goldstein
2025	J	`jnl`	Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs. CoRR Nikita Afonin, Nikita Andriyanov, Nikhil Bageshpura, Kyle Liu, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Alexander Panchenko, Oleg Rogov, Elena Tutubalina, Mikhail Seleznyov
2025	J	`jnl`	Evaluation Awareness Scales Predictably in Open-Weights Large Language Models. CoRR Maheep Chaudhary, Ian Su, Nikhil Hooda, Nishith Shankar, Julia Tan, Kevin Zhu, Ashwinee Panda, Ryan Lagasse, Vasu Sharma
2025	J	`jnl`	FRIT: Using Causal Importance to Improve Chain-of-Thought Faithfulness. CoRR Anand Swaroop, Akshat Nallani, Saksham Uboweja, Adiliia Uzdenova, Michael Nguyen, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Vasu Sharma, Maheep Chaudhary
2025	J	`jnl`	FineGRAIN: Evaluating Failure Modes of Text-to-Image Models with Vision Language Model Judges. CoRR Kevin David Hayes, Micah Goldblum, Vikash Sehwag, Gowthami Somepalli, Ashwinee Panda, Tom Goldstein
2025	J	`jnl`	Gemstones: A Model Suite for Multi-Faceted Scaling Laws. CoRR Sean McLeish, John Kirchenbauer, David Yu Miller, Siddharth Singh, Abhinav Bhatele, Micah Goldblum, Ashwinee Panda, Tom Goldstein
2025	J	`jnl`	LoRI: Reducing Cross-Task Interference in Multi-Task Low-Rank Adaptation. CoRR Juzheng Zhang, Jiacheng You, Ashwinee Panda, Tom Goldstein
2025	J	`jnl`	Modeling and Predicting Multi-Turn Answer Instability in Large Language Models. CoRR Jiahang He, Rishi Ramachandran, Neel Ramachandran, Aryan Katakam, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Aryan Shrivastava
2025	J	`jnl`	Peek-a-Boo Reasoning: Contrastive Region Masking in MLLMs. CoRR Isha Chaturvedi, Anjana Nair, Yushen Li, Adhitya Rajendra Kumar, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Vasu Sharma
2025	A*	`conf`	Privacy Auditing of Large Language Models. ICLR Ashwinee Panda, Xinyu Tang, Christopher A. Choquette-Choo, Milad Nasr, Prateek Mittal
2025	J	`jnl`	Privacy Auditing of Large Language Models. CoRR Ashwinee Panda, Xinyu Tang, Milad Nasr, Christopher A. Choquette-Choo, Prateek Mittal
2025	J	`jnl`	Private Fine-tuning of Large Language Models with Zeroth-order Optimization. Trans. Mach. Learn. Res. Xinyu Tang, Ashwinee Panda, Milad Nasr, Saeed Mahloujifar, Prateek Mittal
2025	J	`jnl`	SALT: Steering Activations towards Leakage-free Thinking in Chain of Thought. CoRR Shourya Batra, Pierce Tillman, Samarth Gaggar, Shashank Kesineni, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Vasu Sharma, Maheep Chaudhary
2025	A*	`conf`	Safety Alignment Should be Made More Than Just a Few Tokens Deep. ICLR Xiangyu Qi, Ashwinee Panda, Kaifeng Lyu, Xiao Ma, Subhrajit Roy, Ahmad Beirami, Prateek Mittal, Peter Henderson
2025	J	`jnl`	Shared Parameter Subspaces and Cross-Task Linearity in Emergently Misaligned Behavior. CoRR Daniel Aarao Reis Arturi, Eric Zhang, Andrew Ansah, Kevin Zhu, Ashwinee Panda, Aishwarya Balwani
2025	J	`jnl`	Using Attention Sinks to Identify and Evaluate Dormant Heads in Pretrained LLMs. CoRR Pedro Sandoval Segura, Xijun Wang, Ashwinee Panda, Micah Goldblum, Ronen Basri, Tom Goldstein, David Jacobs
2024	A*	`conf`	A New Linear Scaling Rule for Private Adaptive Hyperparameter Optimization. ICML Ashwinee Panda, Xinyu Tang, Saeed Mahloujifar, Vikash Sehwag, Prateek Mittal
2024	—	`conf`	Dense Backpropagation Improves Routing for Sparsely-Gated Mixture-of-Experts. ENLSP Ashwinee Panda, Vatsal Baherwani, Zain Sarwar, Benjamin Thérien, Sambit Sahu, Stephen Rawls, Supriyo Chakraborty, Tom Goldstein
2024	J	`jnl`	Lottery Ticket Adaptation: Mitigating Destructive Interference in LLMs. CoRR Ashwinee Panda, Berivan Isik, Xiangyu Qi, Sanmi Koyejo, Tsachy Weissman, Prateek Mittal
2024	A*	`conf`	Privacy-Preserving In-Context Learning for Large Language Models. ICLR Tong Wu, Ashwinee Panda, Jiachen T. Wang, Prateek Mittal
2024	J	`jnl`	Private Fine-tuning of Large Language Models with Zeroth-order Optimization. CoRR Xinyu Tang, Ashwinee Panda, Milad Nasr, Saeed Mahloujifar, Prateek Mittal
2024	J	`jnl`	Refusal Tokens: A Simple Way to Calibrate Refusals in Large Language Models. CoRR Neel Jain, Aditya Shrivastava, Chen Zhu, Daben Liu, Alfy Samuel, Ashwinee Panda, Anoop Kumar, Micah Goldblum, Tom Goldstein
2024	J	`jnl`	Safety Alignment Should Be Made More Than Just a Few Tokens Deep. CoRR Xiangyu Qi, Ashwinee Panda, Kaifeng Lyu, Xiao Ma, Subhrajit Roy, Ahmad Beirami, Prateek Mittal, Peter Henderson
2024	—	`conf`	StructMoE: Structured Mixture of Experts Using Low Rank Experts. ENLSP Zain Sarwar, Ashwinee Panda, Benjamin Thérien, Stephen Rawls, Anirban Das, Kartik Balasubramaniam, Berkcan Kapusuzoglu, Shixiong Zhang, Sambit Sahu, Milind Naphade, Supriyo Chakraborty
2024	A*	`conf`	Teach LLMs to Phish: Stealing Private Information from Language Models. ICLR Ashwinee Panda, Christopher A. Choquette-Choo, Zhengming Zhang, Yaoqing Yang, Prateek Mittal
2024	J	`jnl`	Teach LLMs to Phish: Stealing Private Information from Language Models. CoRR Ashwinee Panda, Christopher A. Choquette-Choo, Zhengming Zhang, Yaoqing Yang, Prateek Mittal
2024	A*	`conf`	Visual Adversarial Examples Jailbreak Aligned Large Language Models. AAAI Xiangyu Qi, Kaixuan Huang, Ashwinee Panda, Peter Henderson, Mengdi Wang, Prateek Mittal
2023	A*	`conf`	Differentially Private Image Classification by Learning Priors from Random Processes. NeurIPS Xinyu Tang, Ashwinee Panda, Vikash Sehwag, Prateek Mittal
2023	J	`jnl`	Differentially Private Image Classification by Learning Priors from Random Processes. CoRR Xinyu Tang, Ashwinee Panda, Vikash Sehwag, Prateek Mittal
2023	J	`jnl`	Differentially Private In-Context Learning. CoRR Ashwinee Panda, Tong Wu, Jiachen T. Wang, Prateek Mittal
2023	J	`jnl`	Visual Adversarial Examples Jailbreak Large Language Models. CoRR Xiangyu Qi, Kaixuan Huang, Ashwinee Panda, Mengdi Wang, Prateek Mittal
2022	J	`jnl`	DP-RAFT: A Differentially Private Recipe for Accelerated Fine-Tuning. CoRR Ashwinee Panda, Xinyu Tang, Vikash Sehwag, Saeed Mahloujifar, Prateek Mittal
2022	A*	`conf`	Neurotoxin: Durable Backdoors in Federated Learning. ICML Zhengming Zhang, Ashwinee Panda, Linyue Song, Yaoqing Yang, Michael W. Mahoney, Prateek Mittal, Kannan Ramchandran, Joseph Gonzalez
2022	J	`jnl`	Neurotoxin: Durable Backdoors in Federated Learning. CoRR Zhengming Zhang, Ashwinee Panda, Linyue Song, Yaoqing Yang, Michael W. Mahoney, Joseph E. Gonzalez, Kannan Ramchandran, Prateek Mittal
2022	A	`conf`	SparseFed: Mitigating Model Poisoning Attacks in Federated Learning with Sparsification. AISTATS Ashwinee Panda, Saeed Mahloujifar, Arjun Nitin Bhagoji, Supriyo Chakraborty, Prateek Mittal
2021	J	`jnl`	SparseFed: Mitigating Model Poisoning Attacks in Federated Learning with Sparsification. CoRR Ashwinee Panda, Saeed Mahloujifar, Arjun Nitin Bhagoji, Supriyo Chakraborty, Prateek Mittal
2020	A*	`conf`	FetchSGD: Communication-Efficient Federated Learning with Sketching. ICML Daniel Rothchild, Ashwinee Panda, Enayat Ullah, Nikita Ivkin, Ion Stoica, Vladimir Braverman, Joseph Gonzalez, Raman Arora
2020	J	`jnl`	FetchSGD: Communication-Efficient Federated Learning with Sketching. CoRR Daniel Rothchild, Ashwinee Panda, Enayat Ullah, Nikita Ivkin, Ion Stoica, Vladimir Braverman, Joseph Gonzalez, Raman Arora