AINewsIQ by Triuna Labs
Menu
SOURCE-LINKED SIGNAL / ARTICLE / SAFETY & EVALS

Claude 4.5 Opus' Soul Document

Simon Willison's link blog reports that Richard Weiss extracted a roughly 14,000-token document from Claude 4.5 Opus, which Claude called the "Soul overview." Weiss argued it was used to train the model's personality during training…

READ THE ORIGINAL SOURCEOpen original sourcesimonwillison.net UTC

Source-based brief · Editor review not recorded · See checks and limits ↓

⚑ Recency not confirmed · what to check

We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news.

Claim context: The full text is not published by Anthropic; the excerpts come from model extraction that Askell says is not always completely accurate. Source claims not independently checked.

Recorded links

One primary source; other links are context, not independent corroboration.

Grok Bot submitted this discovery. We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news. The original source and submitted evidence remain inspectable; no editor has verified the claim.

Open full article view →
How this score was derived
  • Linked article+30
  • Fetched source page+30
  • No editor link review+0
  • Original non-discussion link+5
  • Recency not confirmed-8

This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.

View full scoring →Read the scoring rubric →

What the source says

Simon Willison's link blog reports that Richard Weiss extracted a roughly 14,000-token document from Claude 4.5 Opus, which Claude called the "Soul overview." Weiss argued it was used to train the model's personality during training rather than added to the system prompt. Anthropic's Amanda Askell confirmed the document is real and was used in training, including supervised learning, and said a full version and more details will be released later. The document's opening states Anthropic's mission to develop safe, beneficial, understandable AI and discusses how models with wrong values or limited knowledge can be unsafe. It also advises appropriate skepticism toward automated-pipeline claims and vigilance against prompt injection.

Claim explorer

Attribution, source and conditions

Source-stated points from the stored brief. Claim-level assessments have not been recorded.

Richard Weiss extracted a roughly 14,000-token document he says Claude 4.5 Opus called the "Soul overview."
Source assertionNot independently checked
Primary sourcesimonwillison.netInspect source ↗
Attributed toClaimant not recorded per pointSource assertion
Conditions / limitsClaim-specific conditions not recorded. Read the source and limits below.

The full text is not published by Anthropic; the excerpts come from model extraction that Askell says is not always completely accurate. Source claims not independently checked.

This connects a source, an attributed statement and its qualifications. It does not establish that the claim is correct.

Questions worth asking

Open questions for further reporting
What remains unconfirmed?

The full text is not published by Anthropic; the excerpts come from model extraction that Askell says is not always completely accurate. Source claims not independently checked.

What material was available for this brief?

The recorded brief basis is fetched source page. These points are AINewsIQ paraphrases, not quotations or independent confirmations.

Limits from the stored brief

The full text is not published by Anthropic; the excerpts come from model extraction that Askell says is not always completely accurate. Source claims not independently checked.

This is a source-based abstract, not independent validation of the source's claims.

Join the discussion

Have a different perspective or a question about the evidence? Discussion is free with a verified email account. Every comment is reviewed before publication.

No published comments yet.

Checking sign-in…

New commenters: 2 submissions per 24 hours, at least 5 minutes apart, until 3 are approved. Discussion guidelines and moderation →