WHAT NEW IN X11.04 WED

X 上的 AI 合集档案 · 22 个合集 · 66,612 条

claudeIMAGE

Anthropic 反蒸馏:提取量降 90–99%,快速封号成关键

Anthropic 称新防护让 Claude 数据蒸馏提取量下降 10–100 倍,靠保留思考、两阶段分类器和快速封号,曾一次封掉 2 万多个欺诈账号。

Rich Nicholls
@_richnicholls14 天前收录原推发布于 2026.10.19
Anthropic 反蒸馏:提取量降 90–99%,快速封号成关键
Rich Nicholls @_richnicholls原推发布于 2026.10.19
在 X 上打开原推

New safeguards at @AnthropicAI cut the data distillers' harvest from Claude by about 90 to 99%, early evidence suggests. Measured on its own platform, reasoning tokens extracted fell 10 to 100x after the classifier went in Preserved thinking, a two-stage classifier, and fast account bans are key parts of the stack One network ran more than 20,000 fraudulent accounts at once Zhipu and Moonshot are named as past distillation cases The company says distillation lets rivals mimic frontier capability for a fraction of the development cost. I would count extraction defenses as part of a lab's product, alongside the model itself.

x.com/_richnicholls/status/2112151809620074610采集于 2026.10.20 10:13