Native Robustness

Native robustness in machine learning focuses on developing models inherently resistant to various forms of input perturbations, including adversarial attacks and noisy data, without relying solely on post-hoc defenses. Current research emphasizes techniques like ensemble methods, reprogramming existing models, and modifying training procedures (e.g., using different learning rates for specific model layers or incorporating regularization methods) to improve robustness across diverse model architectures, including convolutional neural networks, vision transformers, and large language models. This field is crucial for deploying reliable AI systems in safety-critical applications, such as healthcare and autonomous driving, where model resilience to unexpected inputs is paramount.

Papers

October 19, 2022

On the Adversarial Robustness of Mixture of Experts
Joan Puigcerver, Rodolphe Jenatton, Carlos Riquelme, Pranjal Awasthi, Srinadh Bhojanapalli
Adversarial Attack Native Robustness Adversarial Robustness Mixture Component Expert Knowledge Better Robustness Dense Model

October 18, 2022

Improving robustness of spontaneous speech synthesis with linguistic speech regularization and pseudo-filled-pause insertion
Yuta Matsunaga, Takaaki Saeki, Shinnosuke Takamichi, Hiroshi Saruwatari
Native Robustness Synthesized Speech Subword Regularization Inappropriate Pause Spontaneous Speech Synthesis

October 16, 2022

Investigating the Robustness of Natural Language Generation from Logical Forms via Counterfactual Samples
Chengyuan Liu, Leilei Gan, Kun Kuang, Fei Wu
Native Robustness Language Generation Counterfactual Reasoning Counterfactual Data Logical Reasoning Ability Logical Form Counterfactual Sample

October 14, 2022

Pretrained Transformers Do not Always Improve Robustness
Swaroop Mishra, Bhavdeep Singh Sachdeva, Chitta Baral
Native Robustness Pre Trained Transformer Robust Representation Natural Robustness Research

October 12, 2022

Double Bubble, Toil and Trouble: Enhancing Certified Robustness through Transitivity
Andrew C. Cullen, Paul Montague, Shijie Liu, Sarah M. Erfani, Benjamin I. P. Rubinstein
Native Robustness Tiny ImageNet Manual Effort Certified Robustness Strong Transitivity Relation Asset Price Bubble Sex Trouble Fast Minimum Norm Attack

October 11, 2022

Improving Robustness of Retrieval Augmented Translation via Shuffling of Suggestions
Cuong Hoang, Devendra Sachan, Prashant Mathur, Brian Thompson, Marcello Federico
Native Robustness Neural Machine Translation Shuffle Model Domain Discrepancy Suggestion System Translation Memory Fuzzy String Matching

October 10, 2022

October 9, 2022

Performances of Symmetric Loss for Private Data from Exponential Mechanism
Jing Bi, Vorapong Suppakitpaisarn
Native Robustness Private Data Private Learning Exponential Mechanism Symmetric Loss

October 7, 2022

October 6, 2022

Towards Out-of-Distribution Adversarial Robustness
Adam Ibrahim, Charles Guille-Escuret, Ioannis Mitliagkas, Irina Rish, David Krueger, Pouya Bashivan
Native Robustness Adversarial Robustness Extrapolation Framework Training Time Attack Domain Generalisation

October 5, 2022

A Closer Look at Robustness to L-infinity and Spatial Perturbations and their Composition
Luke Rowe, Benjamin Thérien, Krzysztof Czarnecki, Hongyang Zhang
Native Robustness Adversarial Learning Glance Annotation Compositional Ability Norm Attack Composite Attack Compositional Adversarial

October 4, 2022

On the Robustness of Deep Clustering Models: Adversarial Attacks and Defenses
Anshuman Chhabra, Ashwin Sekhari, Prasant Mohapatra
Adversarial Attack Native Robustness Deep Clustering

Native Robustness

Papers

On the Adversarial Robustness of Mixture of Experts

Improving robustness of spontaneous speech synthesis with linguistic speech regularization and pseudo-filled-pause insertion

Statistical, Robustness, and Computational Guarantees for Sliced Wasserstein Distances

DE-CROP: Data-efficient Certified Robustness for Pretrained Classifiers

A.I. Robustness: a Human-Centered Perspective on Technological Challenges and Opportunities

Investigating the Robustness of Natural Language Generation from Logical Forms via Counterfactual Samples

Pretrained Transformers Do not Always Improve Robustness

Double Bubble, Toil and Trouble: Enhancing Certified Robustness through Transitivity

Improving Robustness of Retrieval Augmented Translation via Shuffling of Suggestions

Is your noise correction noisy? PLS: Robustness to label noise with two stage detection

Everything is Varied: The Surprising Impact of Individual Variation on ML Robustness in Medicine

Performances of Symmetric Loss for Private Data from Exponential Mechanism

Robustness of Unsupervised Representation Learning without Labels

Enhance Sample Efficiency and Robustness of End-to-end Urban Autonomous Driving via Semantic Masked World Model

ViewFool: Evaluating the Robustness of Visual Recognition to Adversarial Viewpoints

1st ICLR International Workshop on Privacy, Accountability, Interpretability, Robustness, Reasoning on Structured Data (PAIR^2Struct)

A2: Efficient Automated Attacker for Boosting Adversarial Training

Towards Out-of-Distribution Adversarial Robustness

A Closer Look at Robustness to L-infinity and Spatial Perturbations and their Composition

On the Robustness of Deep Clustering Models: Adversarial Attacks and Defenses