Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
AI is already writing a majority of code for frontier AI labs. This creates a safety risk if there is insufficient human oversight. Existing work proposes minimum comprehension thresholds and unaided...
Product-centric advertisement video generation aims to create promotional videos that preserve fine-grained product identity while presenting selling points through coherent multi-shot narratives. How...
Can a biological wiring diagram serve as a useful computational substrate beyond the behaviors for which it evolved? We use the publicly released MaleCNS v1.0 connectome, reconstructed from a single a...
When does distribution alignment help a frozen foundation-model embedding generalize across acoustic domains? For cross-domain mosquito-species classification we report a strength-monotonic law: the s...
The wide adoption of LLMs across broad NLG applications heightens the importance of providing users with the means to avert errors and hallucinations. Uncertainty quantification is poised to fill that...
We study the accuracy of Gauss-Newton curvature in ridge-regularized nonlinear least squares. Under local regularity and persistence of level-set curvature magnitude along an exact-fit section, we pro...
Random selection is widely used in resource allocation and auditing, making reliable implementation essential for AI-agent systems. Behavioural tests across six reasoning models uncovered threshold an...
DP-SGD protects training data by adding Gaussian noise to clipped gradients. The amount of noise is usually chosen by running a numerical privacy accountant inside a search. We study DP-SGD with rando...
The evaluation of coding agents generally strives to be as realistic as possible. In our study, we collect 4,782 agent sessions of real software engineers in JetBrains IDEs, which we call Production S...
Large neural networks often acquire capabilities that small models fail to learn. Does this stem from large models learning more representative features, or from being more robust to unaccounted-for a...
Modern vehicles rely on large numbers of Electronic Control Units (ECUs) that constantly exchange information over the Controller Area Network (CAN) bus. Due to the rapidity, structure, and repetition...
Large language models (LLMs) are increasingly used for emotional support, raising concern that sustained use may draw users away from their real-world relationships. Yet existing evaluations primarily...
Entity Alignment (EA) identifies equivalent entities across knowledge graphs and is critical for knowledge base integration and ontology merging. Evaluating EA systems at scale requires expensive expe...
Open-vocabulary detection (OVD) recognizes categories unseen during training through textual category queries, yet achieving strong generalization with real-time efficiency remains challenging. Beyond...
Spoken language identification (LID) aims to recognize the target language regardless of accent. In practice, however, LID models fine-tuned from self-supervised speech representations frequently conf...
Zero-shot Chinese character recognition (ZS-CCR) aims to recognize characters whose categories are never observed during training, and typically relies on the compositional structure shared between se...
The paper introduces cylindrical geodesic flow matching, a method for translating quasiperiodic cardiovascular waveforms between different body locations. By replacing the standard affine path in flow matching with a closed‑form geodesic on a phase–amplitude cylinder, the approach preserves amplitude and instantaneous frequency during interpolation. Experiments on photoplethysmography and seismocardiography data show that this geometry‑aware method outperforms interpolation baselines and matches or exceeds supervised models, reducing error metrics by up to ~15%.
CausalBind introduces a causal modeling framework for protein‑molecule virtual screening that focuses on sparse cross‑modality interactions rather than dense holistic alignment. By formalizing a V‑structure causal model with Heckman‑style selection, the authors prove that latent interaction concepts are identifiable only under sparsity constraints and propose three implementation variants. Experiments on DUD‑E and LIT‑PCBA benchmarks demonstrate that all variants outperform strong retrieval baselines, especially in early enrichment and out‑of‑distribution generalization.
Local2Mesh is a spatially localized contour‑to‑mesh framework that reconstructs 3‑D left ventricular geometry from sparse 2‑D cardiac MRI contours without requiring 3‑D mesh annotations. It corrects inter‑slice misalignment through geometry‑aware alignment and routes contour features to template vertices via a plane‑aware Local Router, then jointly uses local and global features to deform a template mesh. Experiments on M&Ms‑2 and ACDC datasets show superior geometric reconstruction, functional estimation, and disease classification compared to existing methods, with strong cross‑dataset generalization.
The paper introduces Robust Async-Fed-Q, an epoch‑based federated Q‑learning algorithm that combines variance‑reduced Bellman operator estimation with robust server aggregation. It provides high‑probability finite‑time guarantees showing that honest agents retain sample‑efficiency benefits even when some agents act adversarially, with the adversarial impact diminishing as honest data grows. The authors also present matching lower bounds, extend the framework to single‑trajectory sampling and heterogeneous coverage, and improve communication complexity for asynchronous federated Q‑learning.