Code Reasoning on HumanE
84.9AccuracyDenser
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DenserModel=GPT-4o2025.12 | 84.9 | — | |
| Process SupervisionModel=GPT-4o2025.12 | 84.6 | — | |
| Reflection-CoTModel=GPT-4o2025.12 | 84 | — | |
| Self-VerificationModel=GPT-4o2025.12 | 83.7 | — | |
| Tree-of-ThoughtModel=GPT-4o2025.12 | 83.2 | — | |
| DenserModel=DeepSeek-V32025.12 | 83.1 | — | |
| Process SupervisionModel=DeepSeek-V32025.12 | 82.9 | — | |
| Self-ConsistencyModel=GPT-4o2025.12 | 82.5 | — | |
| Reflection-CoTModel=DeepSeek-V32025.12 | 82.3 | — | |
| Self-VerificationModel=DeepSeek-V32025.12 | 81.9 | — | |
| Think-to-ThinkModel=GPT-4o2025.12 | 81.8 | — | |
| Tree-of-ThoughtModel=DeepSeek-V32025.12 | 81.6 | — | |
| Self-ConsistencyModel=DeepSeek-V32025.12 | 80.9 | — | |
| DenserModel=o1-mini2025.12 | 80.8 | — | |
| Process SupervisionModel=o1-mini2025.12 | 80.7 | — | |
| Think-to-ThinkModel=DeepSeek-V32025.12 | 80.2 | — | |
| Reflection-CoTModel=o1-mini2025.12 | 80.1 | — | |
| Chain-of-ThoughtModel=GPT-4o2025.12 | 79.8 | — | |
| Self-VerificationModel=o1-mini2025.12 | 79.5 | — | |
| Tree-of-ThoughtModel=o1-mini2025.12 | 78.9 | — | |
| Self-ConsistencyModel=o1-mini2025.12 | 78.2 | — | |
| Chain-of-ThoughtModel=DeepSeek-V32025.12 | 78.2 | — | |
| DenserBackbone=Qwen3-14B2025.12 | 78.2 | -58.7 | |
| Think-to-ThinkModel=o1-mini2025.12 | 77.5 | — | |
| Self-VerificationBackbone=Qwen3-14B2025.12 | 76.9 | 142.8 | |
| Process SupervisionBackbone=Qwen3-14B2025.12 | 76.3 | 156.3 | |
| Chain-of-ThoughtModel=o1-mini2025.12 | 75.8 | — | |
| TokenSkip (0.9 ratio)Backbone=Qwen3-14B2025.12 | 75.8 | -53.5 | |
| Reflection-CoTBackbone=Qwen3-14B2025.12 | 75.4 | 127.6 | |
| Tree-of-ThoughtBackbone=Qwen3-14B2025.12 | 75 | 345.7 | |
| Self-ConsistencyBackbone=Qwen3-14B2025.12 | 74.1 | 289.4 | |
| Think-to-ThinkBackbone=Qwen3-14B2025.12 | 73.8 | 187.2 | |
| Chain-of-Thought (CoT)Backbone=Qwen3-14B2025.12 | 72.9 | 0 | |
| BeConciseBackbone=Qwen3-14B2025.12 | 71.5 | -54.2 | |
| AbbreWordsBackbone=Qwen3-14B2025.12 | 70.2 | -50.8 |