Reasoning on Reasoning Domain Arithmetic
67.07ScoreSTEER2ADAPT
Evaluation results
| Method | Links | |
|---|---|---|
| STEER2ADAPTBackbone=Qwen-2.5-7B-Instruct2026.02 | 67.07 | |
| Few-ShotBackbone=Qwen-2.5-7B-Instruct, n=22026.02 | 66.29 | |
| ICLBackbone=Qwen-2.5-7B-Instruct2026.02 | 65.92 | |
| CAABackbone=Qwen-2.5-7B-Instruct2026.02 | 65.91 | |
| REPBackbone=Qwen-2.5-7B-Instruct2026.02 | 65.43 | |
| Direct InferenceBackbone=Qwen-2.5-7B-Instruct2026.02 | 64.98 | |
| Few-ShotBackbone=Qwen-2.5-7B-Instruct, n=12026.02 | 64.78 | |
| STEER2ADAPTBackbone=Llama-3.1-8B-Instruct2026.02 | 61.6 | |
| ICLBackbone=Llama-3.1-8B-Instruct2026.02 | 60.93 | |
| STEER2ADAPTBackbone=Mistral-7B-Instruct-v0.12026.02 | 60.24 | |
| Few-ShotBackbone=Llama-3.1-8B-Instruct, n=12026.02 | 59.89 | |
| Few-ShotBackbone=Llama-3.1-8B-Instruct, n=22026.02 | 59.85 | |
| Direct InferenceBackbone=Llama-3.1-8B-Instruct2026.02 | 59.62 | |
| CAABackbone=Mistral-7B-Instruct-v0.12026.02 | 59.49 | |
| CAABackbone=Llama-3.1-8B-Instruct2026.02 | 59.13 | |
| REPBackbone=Llama-3.1-8B-Instruct2026.02 | 58.74 | |
| Direct InferenceBackbone=Mistral-7B-Instruct-v0.12026.02 | 57.59 | |
| ICLBackbone=Mistral-7B-Instruct-v0.12026.02 | 57.49 | |
| REPBackbone=Mistral-7B-Instruct-v0.12026.02 | 56.72 | |
| Few-ShotBackbone=Mistral-7B-Instruct-v0.1, n=12026.02 | 49.69 | |
| Few-ShotBackbone=Mistral-7B-Instruct-v0.1, n=22026.02 | 49.69 |