arXiv:2607.24555v2 Announce Type: replace-cross Abstract: Serving large language models at long context is bottlenecked by the key-value (KV) cache, which…
Category: cs.AI updates on arXiv.org
AeroDPO: Unleashing Lightweight UAV Navigation with High-Fidelity Perception and Automated Preference Optimization
arXiv:2608.07557v2 Announce Type: replace-cross Abstract: Vision-Language Navigation for Unmanned Aerial Vehicles (UAV-VLN) requires rapid and reactive…
MOSAIC: Masked Outsourcing of Secure AI Computations
arXiv:2607.29221v2 Announce Type: replace-cross Abstract: We address the challenge of securely and efficiently outsourcing AI computations from a trusted…
TabDPT-Turbo: Efficient In-Context Learning for Tabular Prediction
arXiv:2608.01400v2 Announce Type: replace-cross Abstract: Tabular foundation models, driven by in-context learning, have rapidly grown in quality and…
EviPathBench: Benchmarking Evidence Acquisition and Reasoning in Vision-Language Models for Whole-Slide Pathology
arXiv:2607.19261v4 Announce Type: replace-cross Abstract: Whole-slide image (WSI) diagnosis requires identifying diagnostically relevant regions,…
GraphVid: Interactive Graph-Controllable Video Generation
arXiv:2607.21580v2 Announce Type: replace-cross Abstract: Controllable video generation remains challenging due to the difficulty of specifying precise…
Discrete Diffusion Models: A Unified Framework from Tokenization to Generation
arXiv:2607.13431v2 Announce Type: replace-cross Abstract: Discrete denoising diffusion models (DDMs) have recently emerged as a compelling alternative to…
Eluna: An Agentic LLM System for Automating Warehouse Operations with Reasoning and Task Execution
arXiv:2607.08960v2 Announce Type: replace-cross Abstract: Warehouse operations are governed by Standard Operating Procedures (SOPs) that encode complex,…
The Caf\’e in Amsterdam: When the Incumbent Becomes the Oracle
arXiv:2607.13393v3 Announce Type: replace-cross Abstract: A field can reformulate its computations freely exactly where its demand is stated independently…
Co-occurring Associated REtained concepts in Diffusion Unlearning
arXiv:2606.24192v2 Announce Type: replace-cross Abstract: Unlearning has emerged as a key technique to mitigate harmful content generation in diffusion…