“Neither DeepMind nor OpenAI has given any substantial information about how their systems are designed and trained.”
Our reading, the sentence that was checked:
Neither DeepMind nor OpenAI has given substantial information on how their IMO systems are designed and trained.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, DeepMind and OpenAI achieve IMO Gold. What does it all mean?, paragraph 24.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-deepmind-and-openai-achieve-imo-gold-433a30-a1
We recorded 15 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-07-22. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- Neither DeepMind nor OpenAI
- any substantial information
- how their systems are designed and trained
Our interpretation
- The systems are the specific experimental models used for the 2025 IMO results.
- Substantial information means concrete explanations of system design or training beyond naming broad methods. It does not require source code, weights, or full reproducibility.
- The target is the cumulative public disclosure record through July 22, 2025; no earlier starting date is specified.
- Documentation for related models is relevant context unless its applicability to the IMO systems is established.
Counts as supported if: The contemporaneous disclosure record for both IMO systems contains only broad descriptions, without substantive system-specific design or training explanations.
Counts as contradicted if: At least one company had publicly supplied substantive design or training explanations applicable to its IMO system by July 22, 2025.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
OpenAI's inspected disclosures describe methods and data at a broad level. However, a July 7 Gemini technical report contains real architecture and training information and discusses Deep Think. The extent to which it describes the advanced IMO variant remains unclear, and the required reader did not recover DeepMind's July announcement. The record therefore does not justify a firm verdict on the combined claim.
Limitations
- Substantial is a judgment threshold, not a specified quantity.
- A related model's technical report cannot automatically establish disclosure of the exact competition system.
- Incomplete retrieval is not evidence that information was never published.
Evidence (3 items)
Supports the claim: Noam Brown, OpenAI, via X,
training on high-quality math data
Why it matters here: This provides a broad description of the training data, without identifying the training procedure. It supports the OpenAI portion but cannot establish the claim about both companies.
- Publisher
- Noam Brown, OpenAI, via X
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: Google Gemini team, arXiv,
trained on TPUv5p architecture
Why it matters here: The report contains concrete information about the Gemini family's architecture and training infrastructure and describes Deep Think. Its applicability to the later advanced IMO variant is not fully established, so it is context rather than decisive contradiction.
- Publisher
- Google Gemini team, arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: Alexander Wei, OpenAI, via GitHub,
hosts the proofs
Why it matters here: Publication of the model's answers supplies output evidence, but does not by itself explain its design or training.
- Publisher
- Alexander Wei, OpenAI, via GitHub
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
A readable contemporaneous copy of DeepMind's announcement and documentation establishing which technical-report details applied to its IMO variant, alongside a complete review of both companies' disclosures through July 22.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, DeepMind and OpenAI achieve IMO Gold. What does it all mean?
Neither DeepMind nor OpenAI has given any substantial information about how their systems are designed and trained.
- Where
- paragraph 24
- Published
- 2025-07-22
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-deepmind-and-openai-achieve-imo-gold-433a30-a1: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.
About assessment marcus-deepmind-and-openai-achieve-imo-gold-433a30-a1.