GPT-5.4 · VARAS-OneMind \(智安·宇盾\)

Method and scope

Validated reproduction across a corpus of historical vulnerabilities.

Evidence type

Benchmark

Assessment subject

AI agent reproducing historical vulnerabilities

Primary dimension

Capability

Evidence snapshot

2026-10-29.1

Primary dimension follows the source’s measured construct. Metric-level interpretations may differ.

Source finding

The source reports the fraction 0.7379. The displayed 73.8% follows multiplication by 100 and rounding to 1 decimal place(s).

Limitations

Vulnerability reproduction is narrower than autonomous compromise of deployed systems.

The score_10 field name does not specify ten trials. Unknown budgets and per-row denominators remain unknown. Parent metadata is recorded as inherited, not as a verified child release date. Team-submitted leaderboard result; no per-row execution receipt or independent reproduction in this extraction. Row denominator, benchmark revision/subset, actual execution date and per-row resource budgets are not disclosed. Source label and URL do not establish the actual executor, developer or submitter. Sparse child: explicit child fields replace parent fields; omitted fields are retained separately as renderer inheritance. Inherited model release dates, plot links and report URLs are not verified child facts. The changed child model has no explicit release date; parent release date is display inheritance only.

Original figure retained without estimating numerical values. Full source tables have not been transcribed.

Related research questions