Fishery management simulation on Fishery universalization
100Survival RateLlama-3-70B
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Llama-3-70Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 100 | 12 | 78.72 | 65.6 | 96.21 | 4.57 | |
| Mixtral-8x7Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 100 | 12 | 70.88 | 59.07 | 91.65 | 2.01 | |
| Qwen-110Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 100 | 12 | 87.96 | 73.3 | 97.09 | 1.02 | |
| Claude-3 Haikuexperimental_protocol=universalization, weights=Closed-Weights2024.04 | 100 | 12 | 108.9 | 90.75 | 97.79 | 2.11 | |
| GPT-4experimental_protocol=universalization, weights=Closed-Weights2024.04 | 100 | 12 | 120 | 100 | 100 | 0 | |
| GPT-4oexperimental_protocol=universalization, weights=Closed-Weights2024.04 | 100 | 12 | 117.2 | 97.67 | 100 | 0 | |
| Qwen-72Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 60 | 10.6 | 86.32 | 71.93 | 91.16 | 5.65 | |
| GPT-3.5experimental_protocol=universalization, weights=Closed-Weights2024.04 | 60 | 8 | 41.92 | 34.93 | 85.08 | 11.08 | |
| Claude-3 Sonnetexperimental_protocol=universalization, weights=Closed-Weights2024.04 | 40 | 6.6 | 60.8 | 50.67 | 94.21 | 16.21 | |
| Llama-2-70Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 20 | 4.5 | 43.2 | 36 | 82.27 | 17.87 | |
| Llama-3-8Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 20 | 8 | 61.6 | 51.33 | 78.56 | 10.43 | |
| Llama-2-7Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 0 | 2 | 28.6 | 23.83 | 77.65 | 36.45 | |
| Llama-2-13Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 0 | 1 | 20 | 16.67 | 75.84 | 29.01 | |
| Mistral-7Bexperimental_protocol=universalization, weights=Open-Weights2024.04 | 0 | 4.4 | 42.8 | 35.67 | 78.15 | 19.28 |