Temporal Reasoning (periodic) on Synthetic Temporal Dataset (10-fold CV)
3.436Negative Log ScoreQwen2.5-3B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-3BManifold=log_lin2025.10 | 3.436 | 18 | |
| Llama-3.2-3BManifold=log_lin2025.10 | 3.6 | 33 | |
| Llama-3.2-3B-ITManifold=log_lin2025.10 | 3.734 | 46 | |
| gemma-2-2bManifold=log_lin2025.10 | 3.769 | 30 | |
| Qwen2.5-3B-ITManifold=log_lin2025.10 | 3.804 | 32 | |
| Llama-3.1-70B-ITManifold=log_lin2025.10 | 3.858 | 60 | |
| Llama-3.1-8B-ITManifold=log_lin2025.10 | 3.882 | 28 | |
| gemma-2-2b-ITManifold=log_lin2025.10 | 3.929 | 14 |