Investment decision alignment on P1 v1 (test)
7.25Overall MSEInvestAgent
Evaluation results
| Method | Links | ||
|---|---|---|---|
| InvestAgentBackbone=Llama-3.1, Training Stage=Post-InvestAlign2025.07 | 7.25 | -44.52 | |
| InvestAgentBackbone=GPT-3.5, Training Stage=Post-InvestAlign2025.07 | 7.46 | -46.84 | |
| InvestAgentBackbone=Qwen-2, Training Stage=Post-InvestAlign2025.07 | 7.46 | -56.69 | |
| Llama-3.1Training Stage=Pre-SFT2025.07 | 13.07 | — | |
| GPT-3.5Training Stage=Pre-SFT2025.07 | 14.03 | — | |
| Qwen-2Training Stage=Pre-SFT2025.07 | 17.22 | — |