Reasoning benchmark
Graphwalks Parents 128K leaderboard
Graphwalks parents 0-128K. Every model the catalog carries a published Graphwalks Parents 128K value for, ranked by that value.
CategoryReasoning
MeasureLong-context graph reasoning
TasksGraph parent-retrieval tasks
DifficultyAlgorithmic long-context reasoning
Published byIntroducing GPT-5.4 mini and nano
Long-context benchmark for recovering parent relationships inside graph tasks.
No model in the catalog has a published Graphwalks Parents 128K score.
The benchmark is defined by Introducing GPT-5.4 mini and nano, but the catalog carries no value for it yet. An absent value is shown as absent here rather than as a zero.