Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Source trail

LMSYS / DeepSeek / CNBC / Bloomberg

Public MindPattern findings, entities, and graph evidence that cite this source.

Findings

1

All-time hits

1

High value

0

Last seen

2026-04-26

Connected entities

DeepSeek V4 SGLang Miles LMSYS

Related findings

2026-04-26 / HACKER NEWSDeepSeek V4 Preview Launches — SGLang + Miles Deliver First Open-Source Day-0 Verified RL Pipeline for Frontier ModelDeepSeek unveiled V4 Pro and V4 Flash with three architecture innovations: hybrid sparse attention (1M-token context), manifold-constrained hyper-connections (mHC), and native FP4 expert weights. LMSYS published the first open-source stack to serve AND train a frontier model on launch day via SGLang+Miles. Benchmarks: 199 tok/s on B200, 266 tok/s on H200, throughput dropping only 10% from 4K to 900K context tokens. Train-inference alignment diff ~0.02-0.03 at step 0.

Open latest cited source