AI BriefingData from Other Models?
Key point
The Claude Opus 4.7 model card specifies synthetic data generated by other models.
Details
In Anthropic's system card, Claude Opus 4.7 is stated to have been trained not only on public information and public/private datasets, but also on synthetic data generated by other models.
The same phrasing appears for Claude Mythos as well, and it explains that data refinement procedures such as deduplication and classification were used during training.
On the other hand, the point is that this expression does not appear for Opus 4.6, which only has a description centered on web crawling.
The key point is that Anthropic specified the use of synthetic data generated by other models in the new model card, and the post confirms this and examines its significance.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
US Appeals Court Upholds Pentagon Blacklisting of Anthropic as Supply Chain Risk