GPT 无后缀型号 · HealthBench Hard · System-card evaluation; exact dataset version unspecified · 历代记录

按发布时间查看 GPT 无后缀型号 的 2 次发布在 HealthBench Hard · System-card evaluation; exact dataset version unspecified 的成绩、配置与披露。

仅为直接前代间的唯一、有效精确值计算差值,要求单位、指标与明确记录的配置和条件一致。

GPT-5

· 本次发布有 2 条成绩记录,含 2 条精确值(可含历史记录)

GPT-5 · HealthBench Hard 成绩与配置
指标与成绩配置与评测条件出处
Score(raw/length-adjusted未注明)46.2%原文 46.2% · 单位 %
公布 未确认
GPT-5 thinking (system card)
设置
System-card reasoning model; effort is not stated for the cited evaluation.
未知项
Exact reasoning effort is not reported for this system-card evaluation.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
Figure 6; thinking model; score (%); source does not distinguish raw from length-adjusted score.
条件
Keep exact printed value; unknown score adjustment is not mapped onto raw or length-adjusted metrics; direction unknown prevents ranking.
在排名中查看此成绩
Score(raw/length-adjusted未注明)25.5%原文 25.5% · 单位 %
公布 未确认
GPT-5 main
设置
Fast, non-reasoning GPT-5 main model in the system card; exact serving settings are not stated.
未知项
Exact serving configuration is not reported.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
Figure 6; main model; score (%); source does not distinguish raw from length-adjusted score.
条件
Keep exact printed value; unknown score adjustment is not mapped onto raw or length-adjusted metrics; direction unknown prevents ranking.
在排名中查看此成绩

GPT-5.4

· 本次发布有 1 条成绩记录,含 1 条精确值(可含历史记录)

GPT-5.4 · HealthBench Hard 成绩与配置
指标与成绩配置与评测条件出处
Length-adjusted score (0–100)40.1 score原文 40.1% · 单位 score
公布 未确认
GPT-5.4 system-card evaluation · effort unspecified
设置
GPT-5.4 as named in system-card Table 6; reasoning effort and tool settings are not stated.
未知项
Reasoning effort、Tool access, grader details, and run settings
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
HealthBench Hard; exact set/version unspecified. Table 6 reports a single score without a separate raw/adjusted split; stored using the shared primary score metric on a 0–100 scale.
条件
GPT-5.4, as named in the current saved system card; reasoning effort and evaluation run settings are not stated.