49 papers found
Adversarial Attacks on Robotic Vision Language Action Models
arXiv (Cornell University)2025
A Vision for Computational Decarbonization of Societal Infrastructure
IEEE Internet Computing20253 citations
OpenUnlearning: Accelerating LLM Unlearning via Unified Benchmarking of Methods and Metrics
arXiv (Cornell University)2025
OS-Harm: A Benchmark for Measuring Safety of Computer Use Agents
arXiv (Cornell University)20251 citations
Unnatural Languages Are Not Bugs but Features for LLMs
arXiv (Cornell University)2025
Joint Distillation for Fast Likelihood Evaluation and Sampling in Flow-based Models
arXiv (Cornell University)2025
Neural Network Verification with Branch-and-Bound for General Nonlinearities
Lecture notes in computer science20254 citations
Weight Ensembling Improves Reasoning in Language Models
arXiv (Cornell University)2025
Automated Black-box Prompt Engineering for Personalized Text-to-Image Generation
arXiv (Cornell University)20241 citations
Neural Network Verification with Branch-and-Bound for General Nonlinearities
arXiv (Cornell University)20242 citations
Understanding Optimization in Deep Learning with Central Flows
arXiv (Cornell University)2024
Rethinking Distance Metrics for Counterfactual Explainability
arXiv (Cornell University)2024
Predicting the Performance of Foundation Models via Agreement-on-the-Line
arXiv (Cornell University)20241 citations
Rethinking LLM Memorization through the Lens of Adversarial Compression
arXiv (Cornell University)2024
Adaptive Data Optimization: Dynamic Sample Selection with Scaling Laws
arXiv (Cornell University)20241 citations
Massive Activations in Large Language Models
arXiv (Cornell University)20247 citations
TOFU: A Task of Fictitious Unlearning for LLMs
arXiv (Cornell University)20243 citations
Improving Alignment and Robustness with Circuit Breakers
arXiv (Cornell University)20246 citations
One-Step Diffusion Distillation through Score Implicit Matching
arXiv (Cornell University)20241 citations
Bayesian Neural Networks with Domain Knowledge Priors
arXiv (Cornell University)20242 citations
HyperCLIP: Adapting Vision-Language models with Hypernetworks
arXiv (Cornell University)2024
Losses over Labels: Weakly Supervised Learning via Direct Loss Construction
Proceedings of the AAAI Conference on Artificial Intelligence20236 citations
Single Image Backdoor Inversion via Robust Smoothed Classifiers
arXiv (Cornell University)2023
Test-Time Adaptation Induces Stronger Accuracy and Agreement-on-the-Line
arXiv (Cornell University)20231 citations
Low-Variance Gradient Estimation in Unrolled Computation Graphs with ES-Single
arXiv (Cornell University)20231 citations
Code of "Sirens' Whisper: Inaudible Near-Ultrasonic Jailbreaks of Speech-Driven LLMs"
arXiv (Cornell University)2023174 citations
TorchDEQ: A Library for Deep Equilibrium Models
arXiv (Cornell University)20233 citations
Function Approximation for Solving Stackelberg Equilibrium in Large Perfect Information Games
Proceedings of the AAAI Conference on Artificial Intelligence20231 citations
Deep Equilibrium Based Neural Operators for Steady-State PDEs
arXiv (Cornell University)20233 citations
Generative Posterior Networks for Approximately Bayesian Epistemic Uncertainty Estimation
arXiv (Cornell University)2023
Understanding Augmentation-based Self-Supervised Representation Learning via RKHS Approximation and Regression
arXiv (Cornell University)20231 citations
Importance of equivariant and invariant symmetries for fluid flow modeling
arXiv (Cornell University)20231 citations
Localized Text-to-Image Generation for Free via Cross Attention Control
arXiv (Cornell University)20232 citations