Benchmark
Graphwalks BFS >128k
reasoning
spatial_reasoning
long_context
text
A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.
语言EN
满分1
参评模型3
模型排名
| 名次 | 模型 | 机构 | 分数 | 来源 |
|---|---|---|---|---|
| 1 | GPT-4.1 | OpenAI | 19.0 | 来源 ↗ |
| 2 | GPT-4.1 mini | OpenAI | 15.0 | 来源 ↗ |
| 3 | GPT-4.1 nano | OpenAI | 2.9 | 来源 ↗ |