GPT-5.6 Terra · Factual error rate on difficult prompts(Internal error-inducing prompts in the GPT-6.1 Terra comparison article; exact dataset version, sample size and user-flagged provenance unspecified.) 22.3% · 主文证据(ev-gpt61-terra-chart-factuality-gpt56-terra-max)

GPT-5.6 Terra 的 Factual error rate on difficult prompts · Internal error-inducing prompts in the GPT-6.1 Terra comparison article; exact dataset version, sample size and user-flagged provenance unspecified. 官方出处。GPT-5.6 Terra max · 22.3%。见《Introducing GPT-6.1 Terra | OpenAI》Official chart #factuality; exact point aria label; model GPT-5.6 Terra; effort Max; embedded Vega data index 4。证据 ev-gpt61-terra-chart-factuality-gpt56-terra-max。

原文与定位

GPT-5.6 Terra · Factual error rate on difficult prompts(Internal error-inducing prompts in the GPT-6.1 Terra comparison article; exact dataset version, sample size and user-flagged provenance unspecified.),GPT-5.6 Terra max配置,Responses with at least one factual error 22.3%(成绩公布 2026-10-26)。见《Introducing GPT-6.1 Terra | OpenAI》Official chart #factuality; exact point aria label; model GPT-5.6 Terra; effort Max; embedded Vega data index 4。 查看成绩条件

Model: GPT-5.6 Terra; Effort: Max; Cost: $0.25; Answers with any factual error: 22.3%
官方材料
Introducing GPT-6.1 Terra | OpenAI
定位
主文 · Official chart #factuality; exact point aria label; model GPT-5.6 Terra; effort Max; embedded Vega data index 4
材料版本
Official GPT-6.1 Terra comparison article, published 2026-10-26; all six charts reviewed and exact chart points cross-checked against official SVG aria-labels and embedded Vega values (143/143). Factuality set/version and sample size are unspecified.
版本公布
披露时间
· 后补披露
采集时间
历史依据
The official comparison article is dated 2026-10-26, after this release. Its chart values are recorded as later comparisons; no launch-time disclosure is claimed for these points.
引用文本

GPT-5.6 Terra · Factual error rate on difficult prompts(Internal error-inducing prompts in the GPT-6.1 Terra comparison article; exact dataset version, sample size and user-flagged provenance unspecified.),GPT-5.6 Terra max配置,Responses with at least one factual error 22.3%(成绩公布 2026-10-26)。官方出处 Introducing GPT-6.1 Terra | OpenAI(主文,Official GPT-6.1 Terra comparison article, published 2026-10-26; all six charts reviewed and exact chart points cross-checked against official SVG aria-labels and embedded Vega values (143/143). Factuality set/version and sample size are unspecified.)Official chart #factuality; exact point aria label; model GPT-5.6 Terra; effort Max; embedded Vega data index 4,https://openai.com/index/introducing-gpt-6-1-terra/。Benchlog https://benchlog.lekshome.cc/evidence/gpt-5-6-terra/ev-gpt61-terra-chart-factuality-gpt56-terra-max/

关联成绩与条件

GPT-5.6 Terra · Factual error rate on difficult prompts 成绩与配置
指标与成绩配置与评测条件证据关系
Responses with at least one factual error22.3%原文 22.3% · 单位 %
公布
GPT-5.6 Terra max
设置
GPT-5.6 Terra; reasoning effort max, directly labeled in the official chart.
未知项
Sampling and harness settings are not stated for this chart.
工具
工具条件未记录
推理预算
max
框架
未记录
协议
Factual error rate on difficult prompts (lower is better); y metric Answers with any factual error; x axis Cost (USD). Lower is better; article describes an internal benchmark of error-inducing prompts without identifying the user-flagged dataset.
条件
GPT-5.6 Terra; reasoning effort Max; chart cost $0.25.

直接支持该成绩

在排名中查看此成绩