AI News

Unemployment Labs report claims that the weights of six Chinese models, including DeepSeek and Moonshot, carry...

Hacker News TrendingUpdated Nov 13 09:57

AI summary

Will DePue of Unemployment Labs published a forensic report claiming that the public weights of six models—DeepSeek-V4-Flash, DeepSeek-V4-Pro, Kimi K3, Kimi K3.1-Pro, MiniMax-M3, and Ant's Ling-3.1-flash—retain an initialization layout consistent with NVIDIA Blackwell chips (148 or 152 SMs), while earlier models in the same series correspond to Hopper. The method is based on the fact that when PyTorch uses Philox to generate initial weights, the thread count varies with the number of SMs. The author reproduced the retained weights of DeepSeek-V3, V4-Flash, and Kimi K3 on physical H100 and B300 machines and a rented GB200, and claims that all 24 models with publicly disclosed hardware match. The report emphasizes that this evidence only indicates the initialization hardware and cannot determine the chip's location, how it was obtained, or whether export controls were violated; AMD MI300X under a special configuration is the only alternative explanation that has not been fully ruled out.

Why read this

SelectedAI score 80

The report uses initialization fingerprints in the model weights to infer the training hardware and provides a reproducible verification method, by which readers can understand the technical path of export control verification.

Topics

Tags

Body

Only a summary is authorized here. Continue at the validated source link.

Source

Hacker News Trending