XYZ Labs xyzlabs.substack.com

馃敟Hottest馃敟 AI and tech news in Greater China. Best data source to understand China's AI revolution with culture context before Western media catches up.

DeepGEMM: DeepSeek Unveils High-Performance Matrix Multiplication Library on Day 3 of Open Source Week
xyzlabs.substack.com路50w路
Discuss: Substack
DeepSeek Kicks Off Open Source Week with FlashMLA: A Game-Changing GPU Optimization for AI
xyzlabs.substack.com路50w路
Discuss: Substack
Grok 3 Takes on the Riemann Hypothesis: A Potential Nobel-Prize Level Breakthrough
xyzlabs.substack.com路50w路
Discuss: Substack
Open Reasoner Zero: A Breakthrough in AI Training Efficiency Matches DeepSeek with Just 1/30th of Training Steps
xyzlabs.substack.com路50w路
Discuss: Substack
DeepSeek Announces "Open Source Week" Initiative with Full Source Code Release
xyzlabs.substack.com路51w路
Discuss: Substack
Large Language Models Show Concerning Tendency to Flatter Users
xyzlabs.substack.com路51w路
Discuss: Substack
Matrix-Zero: Kunlun's Breakthrough in Spatial Intelligence Brings 3D Generation to New Heights
xyzlabs.substack.com路52w路
Discuss: Substack
Goedel-Prover: Princeton Team Creates World's Most Advanced Formal Reasoning Model
xyzlabs.substack.com路52w路
Discuss: Substack
7B Model Outperform DeepSeek R1: A Breakthrough in Test-Time Scaling
xyzlabs.substack.com路52w路
Discuss: Substack
Berkeley Team Recreates DeepSeek's Success for $4,500: How a 1.5B Model Outperformed o1-preview
xyzlabs.substack.com路52w路
Discuss: Substack
Less is More: SJTU's 817 Samples Challenge Traditional AI Scaling Laws
xyzlabs.substack.com路52w路
Discuss: Substack
How Supernote Captured the High-End E-Paper Market Through Environmental Appeal and Premium DTC Strategy
xyzlabs.substack.com路53w路
Discuss: Substack
Behind Closed Doors: China's AI Pioneers Gather for an Exclusive Discourse on DeepSeek
xyzlabs.substack.com路53w路
Discuss: Substack
Stanford Team Replicates OpenAI's o1 and DeepSeek's R1 for Just $50
xyzlabs.substack.com路53w路
Discuss: Substack
CMU Researchers Make Humanoid Robots Move Like LeBron: ASAP Framework Bridges Simulation-Reality Gap
xyzlabs.substack.com路53w路
Discuss: Substack
DeepSeek Researcher Reveals: R1 Trained in Just 2-3 Weeks, Shows Remarkable Evolution During Chinese New Year
xyzlabs.substack.com路53w路
Discuss: Substack
Why DeepSeek and Other Advanced Models Struggle with Overthinking
xyzlabs.substack.com路53w路
Discuss: Substack
Inside TikTok's Parent Company's AI Code Review System: BitsAI-CR Revealed
xyzlabs.substack.com路53w路
Discuss: Substack
DeepSeek Janus Pro: A Split Personality in AI - Brilliant at Vision, Struggling with Creation
xyzlabs.substack.com路53w路
Discuss: Substack
SiliconFlow Partners with Huawei Cloud to Launch DeepSeek R1 & V3 Inference Services
xyzlabs.substack.com路53w路
Discuss: Substack

Keyboard Shortcuts

Navigation
Next / previous item
j/k
Open post
oorEnter
Preview post
v
Post Actions
Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Recommendations
Add interest / feed
Enter
Not interested
x
Go to
Home
gh
Interests
gi
Feeds
gf
Likes
gl
History
gy
Changelog
gc
Settings
gs
Browse
gb
Search
/
General
Show this help
?
Submit feedback
!
Close modal / unfocus
Esc

Press ? anytime to show this help