MM-CBM adds dual concept bottleneck layers to CLIP to enable interpretable multimodal vision tasks, reporting up to 51.26% average accuracy gains over prior CBMs across four benchmarks.
Concept-monitor: Understand- ing dnn training through individual neurons.arXiv preprint arXiv:2304.13346, 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multimodal Concept Bottleneck Models
MM-CBM adds dual concept bottleneck layers to CLIP to enable interpretable multimodal vision tasks, reporting up to 51.26% average accuracy gains over prior CBMs across four benchmarks.