A new benchmark shows that leading LLMs perform poorly on data structure reasoning tasks, with the top model scoring 0.46 on challenging instances.
$” to ensure a unique structure. The final state is a pre-order traversal collecting edge labels, with child edges visited in lexicographical order and “$
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Can LLMs Reason Structurally? Benchmarking via the Lens of Data Structures
A new benchmark shows that leading LLMs perform poorly on data structure reasoning tasks, with the top model scoring 0.46 on challenging instances.