Irminsul: MLA-Native Position-Independent Caching for Agentic LLM Serving

Get full text

Towards Compute-Aware In-Switch Computing for LLMs Tensor-Parallelism on Multi-GPU Systems

Get full text

Accelerating MoE with Dynamic In-Switch Computing on Multi-GPUs

Get full text

A Scalable Digital Twin Framework for Energy Optimization in Data Centers

Get full text

EdgeServing: Deadline-Aware Multi-DNN Serving at the Edge

Get full text

OpenG2G: A Simulation Platform for AI Datacenter-Grid Runtime Coordination

Get full text

Nitsum: Serving Tiered LLM Requests with Adaptive Tensor Parallelism

Get full text

Low-Latency Out-of-Core ANN Search in High-Dimensional Space

Get full text

Bayes Meets Bernstein at the Meta Level: an Analysis of Fast Rates in Meta-Learning with PAC-Bayes

Get full text

An Extensible and Verifiable Language for Query Rewrite Rules

Get full text

Anatomy of a Query: W5H Dimensions and FAR Patterns for Text-to-SQL Evaluation

Get full text

Patch2Vuln: Agentic Reconstruction of Vulnerabilities from Linux Distribution Binary Patches

Get full text

FedAttr: Towards Privacy-preserving Client-Level Attribution in Federated LLM Fine-tuning

Get full text

On the Security of Research Artifacts

Get full text

PACZero: PAC-Private Fine-Tuning of Language Models via Sign Quantization

Get full text

Privacy by Postprocessing the Discrete Laplace Mechanism

Get full text

Autonomous Adversary: Red-Teaming in the age of LLM

Get full text

Pop Quiz Attack: Black-box Membership Inference Attacks Against Large Language Models

Get full text

Constraining Host-Level Abuse in Self-Hosted Computer-Use Agents via TEE-Backed Isolation

Get full text

Efficiently Escaping Saddle Points in Bilevel Optimization

Get full text