Syllogistic Reasoning on Syllogistic Reason
0.9681BLEUVAE (query_add)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| VAE (query_add)Injection=query_add, Train Setup=prior=False2026.01 | 0.9681 | 92 | |
| VAE (query_add)Injection=query_add, Train Setup=weight=0.52026.01 | 0.9461 | 86 | |
| VAE (kv_mem)Injection=kv_mem, Train Setup=weight=0.52026.01 | 0.9452 | 85 | |
| Qwen2.5-0.5B (Base model)Injection=Base model, Train Setup=Zero-shot2026.01 | 0.9014 | 70 | |
| VAE (kv_add)Injection=kv_add, Train Setup=weight=0.52026.01 | 0.487 | 9 | |
| VAE (kv_add)Injection=kv_add, Train Setup=prior=False2026.01 | 0.4839 | 34 | |
| VAE (kv_mem)Injection=kv_mem, Train Setup=prior=False2026.01 | 0.4672 | 33 | |
| VAE (query_add)Injection=query_add, Train Setup=weight=1.02026.01 | 0.4373 | 33 | |
| Qwen2.5-0.5B (Base model)Injection=Base model, Train Setup=Few-shot (FT)2026.01 | 0.1712 | 0 | |
| Qwen2.5-0.5B (Base model)Injection=Base model, Train Setup=Few-shot2026.01 | 0.1534 | 0 | |
| VAE (kv_add)Injection=kv_add, Train Setup=weight=1.02026.01 | 0.0719 | 0 | |
| VAE (kv_mem)Injection=kv_mem, Train Setup=weight=1.02026.01 | 0 | 0 |