Narrative Text Quality Assessment on Autonomous Generation Stage 1 1.0 (test)
79Agentic Win RateAgentic system
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Agentic systemComparison=Narrative vs VDG, Jury Type=3-model jury, Position Randomized=true2026.04 | 79 | — | |
| Agentic systemComparison=Narrative vs Reconstituted, Jury Type=3-model jury, Position Randomized=true2026.04 | 47 | — | |
| Agentic systemComparison=VDG vs VDG, Jury Type=3-model jury, Position Randomized=true2026.04 | 16 | — | |
| Procedural baselineComparison=VDG vs VDG, Jury Type=3-model jury, Position Randomized=true2026.04 | — | 84 | |
| Procedural baselineComparison=Narrative vs VDG, Jury Type=3-model jury, Position Randomized=true2026.04 | — | 21 | |
| Procedural baselineComparison=Narrative vs Reconstituted, Jury Type=3-model jury, Position Randomized=true2026.04 | — | 53 |