Multimodal LLMs reliably solve many CAPTCHA tasks but can be defended by adding fine-grained localization and implicit counting that drops state-of-the-art success from over 95% to 0%.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
COGNITION: From Evaluation to Defense against Multimodal LLM CAPTCHA Solvers
Multimodal LLMs reliably solve many CAPTCHA tasks but can be defended by adding fine-grained localization and implicit counting that drops state-of-the-art success from over 95% to 0%.