Text Generation on Heuristic Generation (test)
5.03LCTG Error RateMARKERGEN
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MARKERGENBackbone=Llama3.1-70B2025.02 | 5.03 | 3.98 | — | |
| MARKERGENBackbone=Qwen2.5-32B2025.02 | 5.34 | 4.14 | — | |
| MARKERGENBackbone=Llama3.1-8B2025.02 | 6.03 | 4.03 | — | |
| MARKERGENBackbone=Qwen2.5-14B2025.02 | 6.35 | 4 | — | |
| MARKERGENBackbone=Qwen2.5-7B2025.02 | 8.51 | 4.13 | — | |
| ImplicitBackbone=Qwen2.5-32B2025.02 | 12.02 | 3.8 | — | |
| ImplicitBackbone=Qwen2.5-14B2025.02 | 21.34 | 3.8 | — | |
| ImplicitBackbone=Llama3.1-8B2025.02 | 21.91 | 3.72 | — | |
| ImplicitBackbone=Llama3.1-70B2025.02 | 27.89 | 3.74 | — | |
| ImplicitBackbone=Qwen2.5-7B2025.02 | 35.69 | 3.42 | — |