Multi-hop Question Answering on 2WikiMultiHopQA (Token-level F1)
65.9Token F1 ScorePDA-GAM
Evaluation Results
| Method | Links | |
|---|---|---|
| PDA-GAMStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=42026.05 | 65.9 | |
| PDA-GAMStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=32026.05 | 65.6 | |
| PDA-TIESStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=42026.05 | 65.4 | |
| PDA-TIESStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=32026.05 | 65.2 | |
| PDA-AveragingStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=42026.05 | 64.9 | |
| PDA-KnOTSStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=42026.05 | 64.7 | |
| PDA-GAMStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=42026.05 | 64.2 | |
| PDA-TIESStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=42026.05 | 64 | |
| PDA-AveragingStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=42026.05 | 63.8 | |
| PDA-TIESStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=32026.05 | 63.8 | |
| PDA-GAMStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=32026.05 | 63.8 | |
| PDA-KnOTSStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=42026.05 | 63.6 | |
| PDA-GAMStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=22026.05 | 63.5 | |
| PDA-TIESStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=22026.05 | 63.4 | |
| PDA-TIESStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=22026.05 | 62.8 | |
| Search-R1Student Model=Qwen3-8B, Preference Delta Scale=7B over 3B, Number of Preference Deltas (∆)=12026.05 | 62.6 | |
| PDA-GAMStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=22026.05 | 62.4 | |
| Joint ORPOStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=42026.05 | 62 | |
| Qwen3Student Model=Qwen3-8B, Preference Delta Scale=4B over 1.7B, Number of Preference Deltas (∆)=12026.05 | 61.5 | |
| Search-R1Student Model=Tülu3-8B, Preference Delta Scale=7B over 3B, Number of Preference Deltas (∆)=12026.05 | 61.3 | |
| Joint ORPOStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=42026.05 | 60.6 | |
| AceSearcherStudent Model=Qwen3-8B, Preference Delta Scale=7B over 3B, Number of Preference Deltas (∆)=12026.05 | 60.1 | |
| Qwen3Student Model=Tülu3-8B, Preference Delta Scale=4B over 1.7B, Number of Preference Deltas (∆)=12026.05 | 59.2 | |
| Llama-3.2Student Model=Qwen3-8B, Preference Delta Scale=3B over 1B, Number of Preference Deltas (∆)=12026.05 | 58.5 | |
| Student BaselineStudent Model=Qwen3-8B2026.05 | 58.1 | |
| AceSearcherStudent Model=Tülu3-8B, Preference Delta Scale=7B over 3B, Number of Preference Deltas (∆)=12026.05 | 57.4 | |
| Llama-3.2Student Model=Tülu3-8B, Preference Delta Scale=3B over 1B, Number of Preference Deltas (∆)=12026.05 | 55.8 | |
| Student BaselineStudent Model=Tülu3-8B2026.05 | 55.1 | |
| Sequential ORPOStudent Model=Qwen3-8B, Number of Preference Deltas (∆)=22026.05 | 52.2 | |
| Sequential ORPOStudent Model=Tülu3-8B, Number of Preference Deltas (∆)=22026.05 | 50.1 |