📄 文章列表
46 篇文章
AlphaSignal
Google's Gemma 4 Gets 70% Faster Prefill and Fixes Its Laziness Problem | AlphaSignal
AlphaSignal
vLLM's TileRT Integration Hits 618 tok/s by Splitting AI's Two Hardest Jobs | AlphaSignal
TDS
RAG Was Always a Temporary Workaround. What is Next?
AlphaSignal
PyTorch 2.13 Ships 12x Faster Attention and 4x Memory Savings for LLM Training | AlphaSignal
TDS
Tail Control: The Counterintuitive Engineering of Reliable Agentic Workflows
TDS
Persistent Latent Memory for Multi-Hop LLM Agents: How a 6G Handover Paper Closes the Agent Cold-Start
AlphaSignal
Thinking Machines debuts a new type of model: Google’s AI mouse promises better human-AI interaction
AlphaSignal
Claude Opus 4.8 advertises its uncertainty: Dynamic workflows manage Anthropic subagents
AlphaSignal
NVIDIA's DynoSim Screens Thousands of GPU Configs 1,500x Faster Than Real Time | AlphaSignal
AlphaSignal
SGLang Breaks 12,000 Tokens per Second on DeepSeek V4 Pro | AlphaSignal
AlphaSignal
NVIDIA's DynoSim Screens Thousands of GPU Configs 1,500x Faster Than Real Time | AlphaSignal
TDS
The Hot Path Belongs to GBDTs, Agents Own the Cold Path: A Payment-Fraud Benchmark
TDS
A Gentle Introduction to Backtracking
TDS
Prefill Is Compute-Bound. Decode Is Memory-Bound. Why Your GPU Shouldn’t Do Both.
TDS
Agentic AI: How to Save on Tokens
TDS
The Counterintuitive Networking Decisions Behind OpenAI’s 131,000-GPU Training Fabric
TDS
The Infrastructure Behind Making Local LLM Agents Actually Useful
TDS
I Built a C++ Backend So My GPU Would Stop Eating Air
TDS
Prefill Once, Fan Out: KV Snapshot Sharing for Multi-Agent LLM Pipelines
TDS