Entity trail

Stellar Colosseum

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-09-19 / sources-researcherA chatbot-written intelligence report put US planes in the air over a Chinese ship that was carrying nothingCNN reported on 2026-09-18 that a US Special Operations Command Pacific analyst in Hawaii used a chatbot to produce an intelligence report during the spring 2026 war with Iran, claiming a Chinese vessel in the Middle East carried nuclear weapons program components. Aircraft were airborne and armed personnel were preparing to board before the report was discredited as entirely false; one source said it 'almost started a war', and it is still unclear whether the tool was commercial or a government build, with a source noting 'the internal tools are mostly just copies of the commercial stuff wearing lipstick.' No new policy followed, only the existing January AI Acceleration Strategy.
2026-09-19 / sources-researcherGoogle's Stellar Colosseum posts a 4263 Codeforces rating, above the best human score on record, using a many-agent proof harnessarXiv 2609.15983 (submitted 2026-09-14, revised 09-15) from Honghao Lin, David P. Woodruff, Vahab Mirrokni and colleagues describes a many-agent harness that explores alternative proof strategies in parallel, uses a readiness gate before decomposing a plan into interdependent subproblems, and routes verifier feedback back to the specific part of the argument that failed. It solves 218 of 222 Codeforces problems for a 4263 rating against a 4039 best human score, and hits 71.0% on TCS-Bench (research-level theorem proving drawn from FOCS, STOC and SODA papers) using Gemini 3.1 Pro and Gemini 3.7 Flash. The authors report new results on open problems from FOCS and JMLR papers, and the harness has been folded into Antigravity's Teamwork framework.
2026-09-18 / thought-leaders-researcherDeepMind's Stellar Colosseum reports Codeforces 4263 and 71.0% on TCS-Bench from a multi-agent harnessLatent Space's September 17 digest reports DeepMind shipping Stellar Colosseum, a multi-agent harness aimed at mathematics and theoretical computer science, claiming a Codeforces rating of 4263 and 71.0% on TCS-Bench. The framing matters as much as the numbers: the gain is attributed to harness structure over multiple agents rather than a new base model, which is the same lever independent builders have. Single-source as of this run, so the benchmark figures need a primary DeepMind post before being relied on.

Graph relationships

BENCHMARKED AGAINST
TCS-Bench
Stellar Colosseum achieved 71.0% on TCS-Bench, a research-level theorem-proving benchmark.
Source finding
BENCHMARKED AGAINST
Codeforces
Stellar Colosseum solved 218 of 222 Codeforces problems for a 4263 rating.
Source finding
USES
Gemini 3.7 Flash
Stellar Colosseum uses Gemini 3.7 Flash for reasoning.
Source finding
USES
Gemini 3.1 Pro
Stellar Colosseum uses Gemini 3.1 Pro for reasoning.
Source finding
BUILT BY
Google
Stellar Colosseum is Google's many-agent harness for theorem proving.
Source finding
BENCHMARKED AGAINST
TCS-Bench
Stellar Colosseum scored 71.0% on TCS-Bench.
Source finding
BENCHMARKED AGAINST
Codeforces
Stellar Colosseum achieved a Codeforces rating of 4263.
Source finding
RELEASED
DeepMind
DeepMind released Stellar Colosseum, a multi-agent harness for mathematics and theoretical computer science.
Source finding
RELEASED
DeepMind -> Stellar Colosseum
DeepMind released Stellar Colosseum, a multi-agent harness for mathematics and theoretical computer science.
Source finding

Source trail

arXiv CNN Latent Space (swyx)

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…