Today for AI
HOT RADAR
opensourceHEAT 13.0°

AI CURATED HOT STORY · 11/14/2026

Developer Rebuilds DeepSeek V3 in 190 Lines of Pure JAX

1 reports archived1 independent sourcesupdated 11/14/2026, 23:04:15
Synthesis & Latest Updates
1 Sources Cross-Validated

A developer released an open-source project that reimplements the DeepSeek V3 architecture from scratch in roughly 190 lines of pure JAX, covering Multi-head Latent Attention, decoupled RoPE, DeepSeekMoE and auxiliary-loss-free load balancing, and pretrains it on a MacBook Air. With no Flax, Equinox or PyTorch dependency and parameters as plain dicts of arrays, the code is readable top to bottom, making it a high-value reference for understanding frontier MoE architectures.

LATEST/A developer released an open-source project that reimplements the DeepSeek V3 architecture from scratch in roughly 190 lines of pure JAX, covering Multi-head Latent Attention, decoupled RoPE, DeepSeekMoE and auxiliary-loss-free load balancing, and pretrains it on a MacBook Air. With no Flax, Equinox or PyTorch dependency and parameters as plain dicts of arrays, the code is readable top to bottom, making it a high-value reference for understanding frontier MoE architectures.

Hourly heat curve

3 fully observed hours
Now
13.0
Peak
13.511-15
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

Coverage timeline

Total 1 reports · Latest first