MIMIC is a new inversion framework that recovers visual concepts from VLM internal states using joint inversion, feature alignment, and three regularizers.
Grad-cam: Visual explanations from deep networks via gradient-based localization,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MIMIC: Multimodal Inversion for Model Interpretation and Conceptualization
MIMIC is a new inversion framework that recovers visual concepts from VLM internal states using joint inversion, feature alignment, and three regularizers.