model evaluationreadout architecturespublished Nov 14, 2026 · arXiv:2611.15592
Small Heads, Strong Decisions: A Controlled Study of Typed-Decision Models
hallucination detectionFlashAttentionpublished Oct 16, 2026 · arXiv:2610.21483
FlashAttention with Statistics: Low-Overhead Attention Features for Compact Hallucination Detectors
AI training infrastructureGPU power reductionpublished Sep 12, 2026 · arXiv:2609.11542
Characterizing Job Power Elasticity for Power-Flexible AI Training
large language modelskernel optimizationpublished Aug 25, 2026 · arXiv:2608.21836
LLM4LLM: Bridging Kernel Benchmarks and Real Deployment via Closed-Loop Agentic Optimization
multi-agent systemsautonomous optimizationpublished Aug 19, 2026 · arXiv:2608.17071
KernelArc: A Multi-Agent Framework for GPU Kernel Optimization
RoPE scalinglong-context inferencepublished Jul 10, 2026 · arXiv:2607.07740
Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE