Adversarial attacks threaten the reliability of deep neural networks, particularly in black-box settings where transferability is essential. However, existing transfer-based attacks often fail when the target model’s architecture or training diverges from the surrogate, due to decision-boundary variation and representation drift. We introduce CORTA, a…
Reasoning models have demonstrated impressive performance in self-reflection and chain-of-thought reasoning. However, they often produce excessively long outputs, leading to prohibitively large key-value (KV) caches during inference. While chain-of-thought inference significantly improves performance on complex reasoning tasks, it can also lead to reasoning failures when…
This session kicks off micro1’s fall virtual event series “Rewriting the Future of Work” in collaboration with the AI Economy Institute (AIEI). Together, they explore how AI is reshaping global labor markets, creating new pathways for worker mobility, fostering inclusive economic participation, and the role…
Recent Large Reasoning Models (LRMs) have shown substantially improved reasoning capabilities over traditional Large Language Models (LLMs) by incorporating extended thinking processes prior to producing final responses. However, excessively lengthy thinking introduces substantial overhead in terms of token consumption and latency, particularly unnecessary for simple…
Despite perceived tensions between Responsible AI (RAI) and business objectives in large technology companies, RAI efforts still advance thanks to the persistent, often invisible work of passionate advocates who take on RAI work, often in addition to their formal roles. In this paper, we examine…
Graph Neural Networks (GNNs) have achieved strong performance across a range of graph representation learning tasks, yet their adversarial robustness in graph classification remains underexplored compared to node classification. While most existing defenses focus on the message-passing component, this work investigates the overlooked role of…
Understanding cellular responses to chemical interventions is critical to the discovery of effective therapeutics. Because individual biological techniques often measure only one axis of cellular response at a time, high-quality multimodal datasets are needed to unlock a holistic understanding of how cells respond to treatments…
The state of the art in human-centric computer vision achieves high accuracy and robustness across a diverse range of tasks. The most effective models in this domain have billions of parameters, thus requiring extremely large datasets, expensive training regimes, and compute-intensive inference. In this paper,…
Code agents have gained widespread adoption due to their strong code generation capabilities and integration with code interpreters, enabling dynamic execution, debugging, and interactive programming capabilities. While these advancements have streamlined complex workflows, they have also introduced critical safety and security risks. Current static safety…
A Computer-Aided Design (CAD) model encodes an object in two coupled forms: a and its resulting \emph{visible geometric shape}.During iterative design, adjustments to the geometric shape inevitably require synchronized edits to the underlying parametric sequence, called \emph{geometry-driven parametric CAD editing}.The task calls for 1) preserving…
Zeroth-order Optimization (ZO) has received wide attention in machine learning, especially when computing full gradient is expensive or even impossible. Recently, ZO has emerged as an important paradigm for memory-efficient fine-tuning of large language models (LLMs), circumventing the memory overhead of backpropagation. However, existing ZO…
We propose VASA-3D, an audio-driven, single-shot 3D head avatar generator. This research tackles two major challenges: capturing the subtle expression details present in real human faces, and reconstructing an intricate 3D head avatar from a single portrait image. To accurately model expression details, VASA-3D leverages…
