Pairwise Preference Comparison on Pairwise Preference Comparisons 1.5B Scale (test)
0.834Avg Preference ScoreSTACKELBERGGDA FOLLOWER
Evaluation Results
| Method | Links | |
|---|---|---|
| STACKELBERGGDA FOLLOWERBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-0.5B2025.12 | 0.834 | |
| STACKELBERGGDA FOLLOWERBase Model=QWEN2.5-1.5B, Opponent Model=RLOO2025.12 | 0.825 | |
| STACKELBERGGDA FOLLOWERBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-1.5B2025.12 | 0.813 | |
| NASH-MD-PGBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-0.5B2025.12 | 0.812 | |
| NASH-MD-PGBase Model=QWEN2.5-1.5B, Opponent Model=RLOO2025.12 | 0.803 | |
| NASH-MD-PGBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-1.5B2025.12 | 0.791 | |
| STACKELBERGGDA LEADERBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-0.5B2025.12 | 0.729 | |
| STACKELBERGGDA LEADERBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-1.5B2025.12 | 0.707 | |
| STACKELBERGGDA LEADERBase Model=QWEN2.5-1.5B, Opponent Model=RLOO2025.12 | 0.69 | |
| STACKELBERGGDA FOLLOWERBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA LEADER2025.12 | 0.687 | |
| NASH-MD-PGBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA LEADER2025.12 | 0.623 | |
| RLOOBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-0.5B2025.12 | 0.621 | |
| RLOOBase Model=QWEN2.5-1.5B, Opponent Model=QWEN2.5-1.5B2025.12 | 0.599 | |
| QWEN2.5-1.5BOpponent Model=QWEN2.5-0.5B2025.12 | 0.521 | |
| STACKELBERGGDA FOLLOWERBase Model=QWEN2.5-1.5B, Opponent Model=NASH-MD-PG2025.12 | 0.511 | |
| NASH-MD-PGBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA FOLLOWER2025.12 | 0.489 | |
| QWEN2.5-0.5BOpponent Model=QWEN2.5-1.5B2025.12 | 0.479 | |
| QWEN2.5-1.5BOpponent Model=RLOO2025.12 | 0.401 | |
| QWEN2.5-0.5BOpponent Model=RLOO2025.12 | 0.379 | |
| STACKELBERGGDA LEADERBase Model=QWEN2.5-1.5B, Opponent Model=NASH-MD-PG2025.12 | 0.377 | |
| STACKELBERGGDA LEADERBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA FOLLOWER2025.12 | 0.313 | |
| RLOOBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA LEADER2025.12 | 0.31 | |
| QWEN2.5-1.5BOpponent Model=STACKELBERGGDA LEADER2025.12 | 0.293 | |
| QWEN2.5-0.5BOpponent Model=STACKELBERGGDA LEADER2025.12 | 0.271 | |
| QWEN2.5-1.5BOpponent Model=NASH-MD-PG2025.12 | 0.209 | |
| RLOOBase Model=QWEN2.5-1.5B, Opponent Model=NASH-MD-PG2025.12 | 0.197 | |
| QWEN2.5-0.5BOpponent Model=NASH-MD-PG2025.12 | 0.188 | |
| QWEN2.5-1.5BOpponent Model=STACKELBERGGDA FOLLOWER2025.12 | 0.187 | |
| RLOOBase Model=QWEN2.5-1.5B, Opponent Model=STACKELBERGGDA FOLLOWER2025.12 | 0.175 | |
| QWEN2.5-0.5BOpponent Model=STACKELBERGGDA FOLLOWER2025.12 | 0.166 |