A benchmark audit shows that video LLM accuracy and true visual grounding are separable, and that most apparent video understanding comes from frame diversity, not temporal order.
Title resolution pending
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
-
Accuracy Without Grounding: Diagnosing Visual Dependency Dissociation in Video LLM Benchmarks
A benchmark audit shows that video LLM accuracy and true visual grounding are separable, and that most apparent video understanding comes from frame diversity, not temporal order.