{"ID":2883666,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.07833","arxiv_id":"2508.07833","title":"MIMIC: Multimodal Inversion for Model Interpretation and Conceptualization","abstract":"Vision Language Models (VLMs) encode multimodal inputs over large, complex, and difficult-to-interpret architectures, which limit transparency and trust. We propose a Multimodal Inversion for Model Interpretation and Conceptualization (MIMIC) framework that inverts the internal encodings of VLMs. MIMIC uses a joint VLM-based inversion and a feature alignment objective to account for VLM's autoregressive processing. It additionally includes a triplet of regularizers for spatial alignment, natural image smoothness, and semantic realism. We evaluate MIMIC both quantitatively and qualitatively by inverting visual concepts across a range of free-form VLM outputs of varying length. Reported results include both standard visual quality metrics and semantic text-based metrics. To the best of our knowledge, this is the first model inversion approach addressing visual interpretations of VLM concepts.","short_abstract":"Vision Language Models (VLMs) encode multimodal inputs over large, complex, and difficult-to-interpret architectures, which limit transparency and trust. We propose a Multimodal Inversion for Model Interpretation and Conceptualization (MIMIC) framework that inverts the internal encodings of VLMs. MIMIC uses a joint VLM...","url_abs":"https://arxiv.org/abs/2508.07833","url_pdf":"https://arxiv.org/pdf/2508.07833v3","authors":"[\"Animesh Jain\",\"Alexandros Stergiou\"]","published":"2025-08-11T10:36:58Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Language Model\"]","has_code":false}