GraphVerse is a new visual graph reasoning benchmark with structure-preserving image edits and a process-sensitive score, and it shows current multimodal LLMs lag far behind humans, especially on paired-image and NP-hard tasks.
For node mapping, the only perfect match is A0 (C, Pink) with B2 (C, Pink), so mapping A0 -> B2 has cost 0
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
GraphVerse: A Comprehensive Visual Graph Reasoning Benchmark for Multimodal Large Language Models
GraphVerse is a new visual graph reasoning benchmark with structure-preserving image edits and a process-sensitive score, and it shows current multimodal LLMs lag far behind humans, especially on paired-image and NP-hard tasks.