Adversarial cross-model code review.
Adversarial Collaboration: An EDGE Lecture by Daniel Kahneman | Edge.org
[EDITOR'S NOTE: In marking this year's 25th anniversary of Edge, we are presenting original lectures from eminent scientists and other thinkers in the empirical world who are changing the way we think about science and our place in the world.]
Forcing Generative Models to Degenerate Ones: The Power of Data...
Growing applications of large language models (LLMs) trained by a third party raise serious concerns on the security vulnerability of LLMs.It has been demonstrated that malicious actors can...
Get Your Bluesky Personality Analyzed (With a Side of Gentle Roasting) | Bluesky Roast
Discover your Bluesky persona through our AI-powered analysis. Get insights about your posting style, community role, and quirks - all served with a dash of humor.
Get Your Bluesky Personality Analyzed (With a Side of Gentle Roasting) | Bluesky Roast
Discover your Bluesky persona through our AI-powered analysis. Get insights about your posting style, community role, and quirks - all served with a dash of humor.
TRAK: Attributing Model Behavior at Scale
The goal of data attribution is to trace model predictions back to training data. Despite a long line of work towards this goal, existing approaches to data attribution tend to force users to choose between computational tractability and efficacy. That is, computationally tractable methods can struggle with accurately attributing model predictions in non-convex settings (e.g., in the context of deep neural networks), while methods that are effective in such regimes require training thousands of models, which makes them impractical for large models or datasets. In this work, we introduce TRAK (Tracing with the Randomly-projected After Kernel), a data attribution method that is both effective and computationally tractable for large-scale, differentiable models. In particular, by leveraging only a handful of trained models, TRAK can match the performance of attribution methods that require training thousands of models. We demonstrate the utility of TRAK across various modalities and scales: image classifiers trained on ImageNet, vision-language models (CLIP), and language models (BERT and mT5). We provide code for using TRAK (and reproducing our work) at https://github.com/MadryLab/trak .
Adversarial internecting - permablog
How to move data to the Atmosphere and screw over someone else's valuation in the process
Adversarial internecting - permablog
How to move data to the Atmosphere and screw over someone else's valuation in the process
Adversarial internecting - permablog
How to move data to the Atmosphere and screw over someone else's valuation in the process
mixup: Beyond Empirical Risk Minimization
Large deep neural networks are powerful, but exhibit undesirable behaviors such as memorization and sensitivity to adversarial examples. In this work, we propose mixup, a simple learning principle to alleviate these issues. In essence, mixup trains a neural network on convex combinations of pairs of examples and their labels. By doing so, mixup regularizes the neural network to favor simple linear behavior in-between training examples. Our experiments on the ImageNet-2012, CIFAR-10, CIFAR-100, Google commands and UCI datasets show that mixup improves the generalization of state-of-the-art neural network architectures. We also find that mixup reduces the memorization of corrupt labels, increases the robustness to adversarial examples, and stabilizes the training of generative adversarial networks.
Adversarial internecting - permablog
How to move data to the Atmosphere and screw over someone else's valuation in the process
Separating signal from noise in coding evaluations
A new analysis from OpenAI reveals issues in SWE-Bench Pro, a popular coding benchmark, raising concerns about reliability and accuracy in evaluating AI models.
Poisoning Attacks against Support Vector Machines
We investigate a family of poisoning attacks against Support Vector Machines (SVM). Such attacks inject specially crafted training data that increases the SVM's test error. Central to the motivation for these attacks is the fact that most learning algorithms assume that their training data comes from a natural or well-behaved distribution. However, this assumption does not generally hold in security-sensitive settings. As we demonstrate, an intelligent adversary can, to some extent, predict the change of the SVM's decision function due to malicious input and use this ability to construct malicious data. The proposed attack uses a gradient ascent strategy in which the gradient is computed based on properties of the SVM's optimal solution. This method can be kernelized and enables the attack to be constructed in the input space even for non-linear kernels. We experimentally demonstrate that our gradient ascent procedure reliably identifies good local maxima of the non-convex validation error surface, which significantly increases the classifier's test error.
Introducing the Elephant-Goldfish Model plugin: scaling design-driven AI software development
Meet the open source Elephant-Goldfish Model plugin: scaling design-driven AI development with long-horizon context and stateless adversarial reviews.
AI Code Reviews | CodeRabbit | Try for Free.
AI-first pull request reviewer with context-aware feedback, line-by-line code suggestions, and real-time chat.