Paper Acceptance Decision on ICLR 2025 (test)
71.92AccuracyCNPE-7B
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CNPE-7BModel Scale=7B2026.03 | 71.92 | 67.32 | 74.08 | 34.64 | 100 | |
| AIScientist(GPT)Evaluation Approach=pointwise - agents, Agent Provider=GPT2026.03 | 69.72 | 53.01 | 64.49 | 12.46 | 78.79 | |
| DeepReview-14BEvaluation Approach=pointwise - models, Model Scale=14B2026.03 | 68.45 | 62.54 | 66.24 | 25.1 | 82.11 | |
| CycleReviewer-8BEvaluation Approach=pointwise - models, Model Scale=8B2026.03 | 66.09 | 53.38 | 65.25 | 9.33 | 69.69 | |
| DeepReview-7BEvaluation Approach=pointwise - models, Model Scale=7B2026.03 | 64.67 | 54.12 | 59.15 | 9.44 | 69.57 | |
| PairReview(GPT)Evaluation Approach=pairwise / listwise, Agent Provider=GPT2026.03 | 64.35 | 58.51 | 61.3 | 17.01 | 73.87 | |
| PairReview(Gemini)Evaluation Approach=pairwise / listwise, Agent Provider=Gemini2026.03 | 62.46 | 56.3 | 60.54 | 12.61 | 71.27 | |
| PairReview(GLM)Evaluation Approach=pairwise / listwise, Agent Provider=GLM2026.03 | 62.46 | 56.3 | 63.25 | 12.61 | 74.99 | |
| NAIPEvaluation Approach=pairwise / listwise2026.03 | 60.25 | 53.47 | 56.65 | 6.95 | 61.04 | |
| AIScientist(Gemini)Evaluation Approach=pointwise - agents, Agent Provider=Gemini2026.03 | 56.15 | 53.97 | 57 | 10.38 | 60.61 | |
| AgentReview(Gemini)Evaluation Approach=pointwise - agents, Agent Provider=Gemini2026.03 | 53.79 | 52.58 | 56.61 | 9.62 | 58.44 | |
| AgentReview(GPT)Evaluation Approach=pointwise - agents, Agent Provider=GPT2026.03 | 50.79 | 50.03 | 55.06 | 6.28 | 55.6 | |
| AgentReview(GLM)Evaluation Approach=pointwise - agents, Agent Provider=GLM2026.03 | 47 | 46.62 | 64.77 | 12.24 | 63.63 | |
| AIScientist(GLM)Evaluation Approach=pointwise - agents, Agent Provider=GLM2026.03 | 38.01 | 35.16 | 59.12 | 4.79 | 60.22 | |
| SEA-EEvaluation Approach=pointwise - models2026.03 | 37.07 | 33.17 | 56.38 | 5.08 | 50.71 |