Symbolic Reasoning on Coin Flip (test)
98.6AccuracyFew-shot Auto-CoT
Evaluation Results
| Method | Links | |
|---|---|---|
| Few-shot Auto-CoTLLM=GPT-3.5-turbo, Protocol=Few-shot2024.04 | 98.6 | |
| Few-shot Manual-CoTLLM=GPT-3.5-turbo, Protocol=Few-shot2024.04 | 98.2 | |
| DUPLLM=GPT-3.5-turbo, Protocol=Zero-shot2024.04 | 97.6 | |
| Zero-shot PS+LLM=GPT-3.5-turbo, Protocol=Zero-shot2024.04 | 95.4 | |
| Zero-shot CoTLLM=GPT-3.5-turbo, Protocol=Zero-shot2024.04 | 94.4 | |
| Least-to-MostLLM=GPT-3.5-turbo, Protocol=Zero-shot2024.04 | 82.8 |