Quick Test Suggests Opus 5.5 Classifiers Block Kernel Requests for Huawei, Trainium Chips
teortaxesTex · x · 2026-09-23
A developer tested Opus 5.5's safety classifiers by asking it to write flash attention kernels for various AI chips: NVIDIA B300, AMD MI355X and TPU7X requests passed, while Huawei 950DT and Amazon Trainium3 were refused. The poster speculates Anthropic is specifically targeting Chinese hardware, possibly reflecting Dario Amadei's stance on China, and calls for systematic classifier probing. Results are from a quick test and not yet verified.
More from Models
- ChatGPT 3.5 felt as good as 5.6 in memory — like replaying a childhood game — flowersslop · 2026-09-23
- Opus 5.5 vs Sol 6 one-shot coding test: "it is not close" — alvelda · 2026-09-23
- TestingCatalog adds email AI brief as GPT-6 Sol/Luna and Opus 5.5 land — testingcatalog · 2026-09-23
- "Opus 5.5" one-shots bass music in JavaScript — but that model doesn't exist — louisvarge · 2026-09-23
- Opus 5.5 xhigh aces the Boeing 747 benchmark in hands-on test — victormustar · 2026-09-23
- Early hands-on: GPT-6-Sol xhigh underwhelms on Codex's Boeing bench — victormustar · 2026-09-23