GPT-6 Astra · 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.) 9.7% · 主文证据(ev-astra-terra61-comparison-factuality-medium)
GPT-6 Astra 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-6 Astra medium · 9.7%。见《Introducing GPT-6.1 Terra — later GPT-6 Astra comparison points》Article chart #factuality / Factual error rate on difficult prompts (lower is better) / GPT-6 Astra / Medium / vegaLiteSpec.data.values[6] and rendered SVG aria-label。证据 ev-astra-terra61-comparison-factuality-medium。
原文与定位
GPT-6 Astra · 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-6 Astra medium配置,Responses with at least one factual error 9.7%(具体成绩公布时间未确认)。见《Introducing GPT-6.1 Terra — later GPT-6 Astra comparison points》Article chart #factuality / Factual error rate on difficult prompts (lower is better) / GPT-6 Astra / Medium / vegaLiteSpec.data.values[6] and rendered SVG aria-label。 查看成绩条件
Factual error rate on difficult prompts (lower is better) — Model: GPT-6 Astra; Effort: Medium; Cost: $0.29; Answers with any factual error: 9.7%; original score=0.097; usd=0.29. Chart y-axis and score tooltip use percent formatting.
- 官方材料
- Introducing GPT-6.1 Terra — later GPT-6 Astra comparison points
- 定位
- 主文 · Article chart #factuality / Factual error rate on difficult prompts (lower is better) / GPT-6 Astra / Medium / vegaLiteSpec.data.values[6] and rendered SVG aria-label
- 材料版本
- Current official article and six rendered Vega charts collected 2026-11-16; article introduces GPT-6.1 Terra on 2026-10-26, exact revision time unknown.
- 版本公布
- 未确认
- 披露时间
- 未确认 · 后补披露
- 采集时间
- 历史依据
- 暂无可靠历史版本依据
引用文本
GPT-6 Astra · 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-6 Astra medium配置,Responses with at least one factual error 9.7%(具体成绩公布时间未确认)。官方出处 Introducing GPT-6.1 Terra — later GPT-6 Astra comparison points(主文,Current official article and six rendered Vega charts collected 2026-11-16; article introduces GPT-6.1 Terra on 2026-10-26, exact revision time unknown.)Article chart #factuality / Factual error rate on difficult prompts (lower is better) / GPT-6 Astra / Medium / vegaLiteSpec.data.values[6] and rendered SVG aria-label,https://openai.com/index/introducing-gpt-6-1-terra/。Benchlog https://benchlog.lekshome.cc/evidence/gpt-6-astra/ev-astra-terra61-comparison-factuality-medium/
关联成绩与条件
| 指标与成绩 | 配置与评测条件 | 证据关系 |
|---|---|---|
| Responses with at least one factual error9.7%原文 9.7% · 单位 % 公布 未确认 | GPT-6 Astra medium
| 直接支持该成绩 在排名中查看此成绩 |