F671 — Openai Real World Reasoning Corpus
F671 covers “Openai Real World Reasoning Corpus”. The canonical record classifies it as candidate/partial, based on 4 source artifacts and 2 original source figures. The highlighted measurements are Direct Recall Total = 30, Pass = No, and Direct Recall Passed = 30.
Purpose and research question
The objective is to document the recorded outcome for “Openai Real World Reasoning Corpus”. The research question is whether the phase-specific checks support the canonical candidate/partial classification. The recorded measurements include Direct Recall Total = 30, Pass = No, and Direct Recall Passed = 30.
Method
F671 is documented by 2 structured JSON records and 2 Markdown reports and 2 original source figures. The reported table retains safe scalar measurements found in that phase package.
Documentation basis
4 source artifacts and 2 phase-specific figures are recorded for this phase.
Evidence highlights
- Direct Recall Total: 30
- Pass: No
- Direct Recall Passed: 30
- Direct Recall Score: 1
- Unknown Defer Passed: Yes
- Anti Forgetting Passed: 3
- Anti Forgetting Score: 0.75
- Chat Regression Passed: 4
Recorded evidence
| Direct Recall Total | 30 |
|---|---|
| Pass | No |
| Direct Recall Passed | 30 |
| Direct Recall Score | 1 |
| Unknown Defer Passed | Yes |
| Anti Forgetting Passed | 3 |
| Anti Forgetting Score | 0.75 |
| Chat Regression Passed | 4 |
Public evidence and source figures
Original source figure · 1920 × 992 px
Original source figure · 1920 × 992 px
These are original source figures from the existing public evidence set, not editorial HTML visualizations.
Result and interpretation
F671 retains its candidate/partial classification and the recorded decision PARTIAL. The recorded measurements include Direct Recall Total = 30, Pass = No, Direct Recall Passed = 30, and Direct Recall Score = 1.
Key results
- Direct Recall Total
- 30
- Pass
- No
- Direct Recall Passed
- 30
- Direct Recall Score
- 1
- Unknown Defer Passed
- Yes
- Anti Forgetting Passed
- 3
- Anti Forgetting Score
- 0.75
- Chat Regression Passed
- 4
Evidence sources
Values on this page are reproduced from the archived phase record and the source artifacts below. No new benchmark was run to create this page.
- OPENAI REAL WORLD REASONING CORPUS 671 (JSON)
- OPENAI REAL WORLD REASONING CORPUS 671 (MD)
- OPENAI REAL WORLD REASONING TRAINING 671 (JSON)
- OPENAI REAL WORLD REASONING TRAINING 671 (MD)