Data Card Generation on MetaGAI-Bench Data Cards
1Faithfulness RankAdaQE-CG
Evaluation Results
| Method | Links | ||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AdaQE-CGLLM Judge=GPT-5-nano2026.03 | 1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 1 | 1 | 1 | 1 | 1 | |
| AdaQE-CGLLM Judge=Gemini-2.5 Flash-Lite2026.03 | 1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 1 | 1 | 1 | 1 | 1 | |
| Zero-shotLLM Judge=Gemini-2.5 Flash-Lite2026.03 | 2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 2 | 2 | 2 | 2 | 2 | |
| Human-AnnotatedLLM Judge=GPT-5-nano2026.03 | 2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 2 | 3 | 3 | 2 | 2.4 | |
| Zero-shotLLM Judge=GPT-5-nano2026.03 | 3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3 | 2 | 2 | 3 | 2.6 | |
| Human-AnnotatedLLM Judge=Gemini-2.5 Flash-Lite2026.03 | 3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3 | 3 | 3 | 3 | 3 | |
| CardGenLLM Judge=GPT-5-nano2026.03 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 4 | 4 | 4 | 4 | 4 | |
| CardGenLLM Judge=Gemini-2.5 Flash-Lite2026.03 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 4 | 4 | 4 | 4 | 4 | |
| AdaQE-CGCard Type=Data2026.03 | — | 2 | 3.39 | 2 | 3.15 | 2 | 3.71 | 2 | 2.83 | 2 | 3.31 | 2 | 2.6 | 2 | 3.32 | 2 | 2.7 | 2 | 2.95 | 2 | 2.42 | 2 | 3.34 | 2 | 2.74 | — | — | — | — | — | |
| CardGenCard Type=Data2026.03 | — | 3 | 3.08 | 3 | 2.98 | 3 | 3.5 | 3 | 2.61 | 3 | 3.08 | 3 | 2.25 | 3 | 3.13 | 3 | 2.53 | 3 | 2.69 | 4 | 2.21 | 3 | 3.1 | 3.2 | 2.52 | — | — | — | — | — | |
| Human-AnnotatedCard Type=Data2026.03 | — | 1 | 3.79 | 1 | 4.01 | 1 | 3.95 | 1 | 3.92 | 1 | 3.57 | 1 | 3.57 | 1 | 3.43 | 1 | 3.56 | 1 | 3.22 | 1 | 3.45 | 1 | 3.59 | 1 | 3.7 | — | — | — | — | — | |
| Zero-shotCard Type=Data2026.03 | — | 4 | 2.94 | 4 | 2.76 | 4 | 3.44 | 4 | 2.41 | 4 | 3.01 | 4 | 2.13 | 4 | 2.95 | 4 | 2.2 | 4 | 2.5 | 3 | 2.23 | 4 | 2.97 | 3.8 | 2.35 | — | — | — | — | — |