The paper introduces AstroVLBench, a benchmark of 4,100 expert‑verified instances covering optical imaging, radio interferometry, multi‑wavelength photometry, time‑domain light curves, and optical spectroscopy. Six leading vision‑language models were evaluated, revealing that performance varies strongly with modality and is not predicted by general benchmark rankings. Mechanistic ablations identified three bottlenecks: prompt grounding depends on model and task, presenting one‑dimensional data as tables rather than plots improves light‑curve classification, and a significant fraction of correct predictions are supported by physically inconsistent reasoning.
By Wenke Ren, Hengxiao Guo, Wenwen Zuo, Xiaoman Zhang
arXiv:2608. 05238v1 Announce Type: new Abstract: Training multimodal models to align time series with language runs into a self-supervision trap.
By Xinran Feng, Yi Xie, Chao Zhang, Ruikun Li, Wanyun Ling, Ziyue Li, Chenxi Liu
The paper introduces D3-Omni, a balanced and decoupled benchmark designed to diagnose fine‑grained multimodal understanding in OmniJudges that evaluate text‑to‑image, text‑to‑video, and text‑to‑speech generation. D3-Omni covers 53 orthogonal binary dimensions across 10,671 samples, using fixed positive seeds and controlled prompt rewriting to generate negatives, thereby ensuring each error can be attributed to a single capability. The benchmark’s dual‑balanced, decoupled, and dynamic design achieves near 1:1 per‑dimension parity and a uniform total‑score distribution, revealing that strong OmniJudges often miss modality‑related failures and treat distinct attributes as a single decision, masking systematic blind spots.
By Guangzheng Hu, Ziyue Jiang, Weixu Qiao, Lixin Zhang, Jianye Kang, Yuru Wu, Rong Bao, Niantong Li, Wei Wang, Ziyi Cheng, Xinfa Zhu, HangRui Hu, Ting He, Bing Zhao, Lin Qu, Hu Wei, Jin Xu
arXiv:2607. 05438v1 Announce Type: cross Abstract: Multimodal retrieval-augmented generation (RAG) grounds a generator in evidence drawn from heterogeneous modalities -- text, tables, and images.
By Xue Li, Yiming Gai
arXiv:2603. 25112v2 Announce Type: replace-cross Abstract: Standard evaluation of LLM confidence relies on calibration metrics (ECE, Brier score) that conflate how much a model knows (Type-1 accuracy) with how well its confidence signal tracks that knowledge (Type-2 metacognitive sensitivity).
By Jon-Paul Cacioli
The paper investigates how new concepts can be integrated into unified multimodal models (UMMs) by separating generation and understanding objectives through a novel visual entity bound to a single task direction. Experiments show that the effectiveness of cross‑task usability depends on where the concept is injected into the shared computation, with a mid‑stack alignment objective achieving high concept acquisition with minimal loss to overall performance. The study highlights that unified weights alone are insufficient; the two directions must share a semantic format at the entry point for efficient concept integration.
By Zongyang Qiu, Yihan Wu, Kaixuan Fan, Bo Li, Hui Xiong