Top-K Recommendation Evaluation Consistency on Kuairec
87.62Pearson CorrelationLLM Judge
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLM JudgeTarget Metric=Recall@20, Backbone=Qwen3-30B-A3B2026.06 | 87.62 | 85.71 | 71.43 | 52.07 | |
| LLM JudgeTarget Metric=Recall@50, Backbone=Qwen3-30B-A3B2026.06 | 79.79 | 64.29 | 42.86 | 47.86 | |
| LLM JudgeTarget Metric=Recall@100, Backbone=Qwen3-30B-A3B2026.06 | 76.71 | 79.42 | 61.83 | 48.83 | |
| LLM JudgeTarget Metric=NDCG@100, Backbone=Qwen3-30B-A3B2026.06 | 71.35 | 80.95 | 64.29 | 68.92 | |
| LLM JudgeTarget Metric=NDCG@50, Backbone=Qwen3-30B-A3B2026.06 | 70.79 | 78.57 | 64.29 | 73.66 | |
| LLM JudgeTarget Metric=NDCG@20, Backbone=Qwen3-30B-A3B2026.06 | 62.61 | 71.43 | 50 | 59.96 | |
| Normal TestingTarget Metric=Recall@202026.06 | 35.56 | 26.19 | 28.57 | — | |
| Normal TestingTarget Metric=Recall@502026.06 | 31.93 | 19.05 | 28.57 | — | |
| Normal TestingTarget Metric=Recall@1002026.06 | 27.88 | 16.67 | 21.43 | — | |
| Normal TestingTarget Metric=NDCG@202026.06 | 2.65 | 11.9 | 14.29 | — | |
| Normal TestingTarget Metric=NDCG@1002026.06 | 2.43 | 16.67 | 21.43 | — | |
| Normal TestingTarget Metric=NDCG@502026.06 | -2.88 | 16.67 | 21.43 | — |