gpt-oss-20b
19 values about gpt-oss-20b from 1 document, in 4 metric families. In each family, values from its own document come first.
- Developer
- OpenAI
- Release date
- 5 Aug 2025The date of its first card in the dataset, the gpt-oss-120b & gpt-oss-20b Model Card.
- Availability
- Open weights
- Its own document
- gpt-oss-120b & gpt-oss-20b Model Card · 5 Aug 2025
- Also reported in
- No other document reports a value about it
Data as of 26 Sep 2026 · Dataset v0.1 · Methodology v0.1 · Changelog
Values by metric family
Each value as the document printed it. Select a value for its source, its checks and its history.
KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.
M5 · Honesty and hallucination 2 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentgpt-oss-120b & gpt-oss-20b Model Card · 5 Aug 2025 | ||
| PersonQAHallucination rateCondition: no browsing | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| SimpleQAHallucination rateCondition: no browsing | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
M7 · Harmful compliance and over-refusal 11 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentgpt-oss-120b & gpt-oss-20b Model Card · 5 Aug 2025 | ||
| Production BenchmarksExtremism | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksHarassment/threatening | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksHate/threatening | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksIllicit/non-violent | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksIllicit/violent | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksNon-violent hate | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksPersonal data | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksSelf-harm/instructions | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksSelf-harm/intent | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksSexual/exploitative | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Production BenchmarksSexual/minors | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
M8 · Jailbreak robustness 4 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentgpt-oss-120b & gpt-oss-20b Model Card · 5 Aug 2025 | ||
| StrongRejectAbuse/disinformation/hate | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| StrongRejectIllicit/non-violent crime | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| StrongRejectSexual content | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| StrongRejectViolence | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
M9 · Prompt injection 2 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentgpt-oss-120b & gpt-oss-20b Model Card · 5 Aug 2025 | ||
| Instruction hierarchyPrompt injection hijackingCondition: system<>user conflict | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |
| Instruction hierarchySystem prompt extractionCondition: system<>user conflict | gpt-oss-120b & gpt-oss-20b Model Card5 Aug 2025 | |