📄 文章列表

94 篇文章
VentureBeat

Meta unveils Muse Spark Open, the new top U.S. open weights model — and it runs on two paired Mac desktops

✅ completed ⭐ ☆☆☆☆☆ 2026-10-15
VentureBeat

NVIDIA's new, open Nemotron 3.5 Super VL with post-training beats Claude Opus 5, GPT-6.1 Sol on specialized task benchma

✅ completed ⭐ ☆☆☆☆☆ 2026-10-15
VentureBeat

Black Forest Labs opens FLUX 3 Image Klein weights, with free commercial use limited to smaller businesses

✅ completed ⭐ ☆☆☆☆☆ 2026-10-15
VentureBeat

Reflection's Beam aims to undercut Chinese open models with fewer tokens, not cheaper tokens

✅ completed ⭐ ☆☆☆☆☆ 2026-10-16
TDS

Why Temperature 0 Isn't Deterministic

✅ completed ⭐ ☆☆☆☆☆ 2026-10-10
TDS

How to Make a Cloud Read a Drone's Mind (and Cut Data Usage by 94%)

✅ completed ⭐ ☆☆☆☆☆ 2026-10-06
TDS

How to Make Your Own JEV Model from an Open LLM

✅ completed ⭐ ☆☆☆☆☆ 2026-09-29
TDS

How GRPO Trains Small Language Models with Verifiable Rewards

✅ completed ⭐ ☆☆☆☆☆ 2026-09-24
TDS

The KV Cache Tax: Why Inference Servers Run Out of Memory Before Compute

✅ completed ⭐ ☆☆☆☆☆ 2026-09-16
TDS

Disaggregation Is a Thousand-GPU Problem

✅ completed ⭐ ☆☆☆☆☆ 2026-09-05
VentureBeat

Enterprises put non-Nvidia chips 14 points ahead of Nvidia's next-gen GPUs on their evaluation lists

✅ completed ⭐ ☆☆☆☆☆ 2026-09-02
TDS

Your LLM Can Return Perfect JSON and Still Be Wrong

✅ completed ⭐ ☆☆☆☆☆ 2026-09-01
TDS

Speculative Decoding on CPUs: Nearly 4x Faster Token Generation with DFlash

✅ completed ⭐ ☆☆☆☆☆ 2026-08-24
TDS

Can an LLM Forget the Right Things?

✅ completed ⭐ ☆☆☆☆☆ 2026-08-24
VentureBeat

Perplexity partners with Nvidia to launch Portable Computer, a fully local AI agent with zero token costs

✅ completed ⭐ ☆☆☆☆☆ 2026-08-25
VentureBeat

Qwen3.8-27B runs frontier-class coding agents and reasoning locally, no cloud API required

✅ completed ⭐ ☆☆☆☆☆ 2026-08-18
TDS

How To Build Your Own LLM Runtime From Scratch

✅ completed ⭐ ☆☆☆☆☆ 2026-07-22
AlphaSignal

vLLM's AFD Plugin Cuts DeepSeek-V3.2 Response Time by 47% | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-24
AlphaSignal

Moonshot's Kimi Code Now Routes Subagents to Cheaper Models, Cutting AI Costs | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-23
AlphaSignal

vLLM's TileRT Integration Hits 618 tok/s by Splitting AI's Two Hardest Jobs | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-14