cs updates on arXiv.org · Scour

TraceCompiler: Skill-Guided Mining and Compilation of LLM Agent Traces into Mostly Deterministic Workflows

SAGE: Semantic Explainability of Attention-Based Survival Models in Computational Pathology

Fast Object Removal Attacks on Safety-Critical Video-based Perception Systems

Towards Designing for (Dis)Trust in Technologies for Aging

Light-Loco-Parkour: Versatile Perceptive Whole-Body Locomotion via Multi-Skill Distillation

Stylometric Defenses Against Author Impersonation in Software Repositories

PolicyGuard: Prompt-Configurable Semantic DLP for LLM Coding Agents

Vulnerabilities, Secrets and Misconfiguration in the Highest-Exposure Docker Hub Images

Better, Stronger, Faster, and Broader: Structured All-Mask Prediction for MLLM-Based Segmentation

Evading Chain-of-Thought Monitoring Through Model Poisoning

Discussed on Hacker News

CT-HEG: A Bidirectional, Timestamp-Attributed Event Graph for ICU In-Hospital Mortality Prediction - An Architectural Ablation Study

Studying, Identifying, and Fixing Hidden Technical Debt in AI-Intensive Cyber-Physical Systems

Stabilization of First-Order Partial Integro-Differential Equations with Concurrent Input and State Delays

Implicit-explicit and split-explicit super-time-stepping methods

Extracting ODRL Policies from Business Process Models: A Graph Traversal Approach to Compliance-by-Extraction

Coordinated Primary Frequency Regulation and Grid-Forming Control for Wind Turbine Generators

MutMem: Cryptographically Authorized Mutation in Persistent Agent Memory

Axiomatic shared-medium coordination for stigmergic systems

KernelBrain: Coarse-to-Fine, Budget-Aware Search for Agentic GPU Kernel Optimization

What the Detector Can See: Evaluating CPS Anomaly Detectors Independently of the Decision Rule

Log in to enable infinite scrolling