Concurrent Negotiation on 128-product In-distribution (test)
0.58RewardQwen3-30B-A3B-Instruct-2507-trained
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen3-30B-A3B-Instruct-2507-trainedParams=30B2026.07 | 0.58 | 70 | 82.8 | 2.47 | 99.5 | 0 | |
| Llama-3.3-70B-InstructParams=70B2026.07 | 0.468 | 16.2 | 36.3 | 2.53 | 48.4 | 35.9 | |
| Qwen3-30B-A3B-nothinkParams=30B2026.07 | 0.386 | 5.5 | 16 | 2.22 | 51.2 | 38.3 | |
| Qwen3-4B-Instruct-2507Params=4B2026.07 | 0.27 | 8 | 9.8 | 2.35 | 52 | 21.9 | |
| Qwen3-30B-A3B-Instruct-2507-untrainedParams=30B2026.07 | 0.151 | 4.8 | 35.5 | 2.39 | 57.1 | 14.8 | |
| gpt-5.1-high-reasoningParams=closed-source2026.07 | 0.123 | 18.1 | 68 | 2.94 | 38.2 | 0 | |
| Kimi-K2.6-nothinkParams=1T2026.07 | 0.101 | 12 | 49.2 | 2.54 | 39.7 | 7.4 | |
| Qwen3-235B-A22B-Instruct-2507Params=235B2026.07 | 0.099 | 11.6 | 56.2 | 1.96 | 44.4 | 5.9 | |
| Kimi-K2.6-thinkingParams=1T2026.07 | 0.077 | 12.8 | 63.8 | 2.2 | 48.4 | 0.4 | |
| gpt-5.1-no-reasoningParams=closed-source2026.07 | 0.046 | 11.5 | 65.3 | 1.74 | 42.9 | 0.4 | |
| DeepSeek-V3.1-nothinkParams=671B2026.07 | 0.041 | 19.2 | 53.9 | 2.38 | 35.5 | 3.9 | |
| Qwen3-30B-A3B-thinkParams=30B2026.07 | 0.021 | 6.2 | 35.9 | 2.46 | 44.6 | 3.9 | |
| DeepSeek-V3.1-thinkingParams=671B2026.07 | 0.016 | 14.8 | 50.8 | 2.62 | 34.6 | 3.1 |