In vision-language models (VLMs), the ability to perceive and interpret color and physical environment is crucial for achieving contextually accurate understanding and interaction. However, despite advances in multimodal modeling, there remains a significant lack of specialized datasets that rigorously evaluate a model's capacity to discern subtle color variations and spatial context---critical elements for situational comprehension and reliable deployment across real-world applications. Toward that goal, we curate MegaCoin, a high-quality, human-labeled dataset based on \emph{real} images with various contextual attributes. MegaCoin consists of two parts: MegaCoin-Instruct, which serves as a supervised fine-tuning (SFT) dataset for VLMs; and MegaCoin-Bench, an annotated test set that can be used as a stand-alone QA dataset. MegaCoin provides three annotated features for 220,000 real images: foreground color, background color, and description of an object's physical environment, constituting 660k human annotations. In addition, MegaCoin can be applied to benchmark domain generalization (DG) algorithms. We explore benchmarking DG methods in the linear probing setup for VLM and show some new insights. Last but not least, we show that VLMs, including GPT-4o, have subpar color recognition capabilities, and fine-tuning with MegaCoin can result in improved performance on visual evaluation tasks. In certain cases, MegaCoin fine-tuned small-scale open-source models such as LLaVA and Bunny can outperform closed-source GPT-4o. We hope the utilities of MegaCoin can shed light on the directions VLMs can improve and provide a more complex platform for domain generalization algorithms.

We present MIDDAG, an intuitive, interactive system that visualizes the information propagation paths on social media triggered by COVID-19-related news articles accompanied by comprehensive insights including user/community susceptibility level, as well as events and popular opinions raised by the crowd while propagating the information. Besides discovering information flow patterns among users, we construct communities among users and develop the propagation forecasting capability, enabling tracing and understanding of how information is disseminated at a higher level. A demo video and more are available at https://info-pathways.github.io.

Conditional nonlinear optimal perturbation (CNOP) is an initial perturbation evolving into the largest nonlinear evolution at the prediction time. It has become a useful tool in meteorology and oceanography. The common method for solving the CNOP is the adjoint-based method which is always referred to as the benchmark. Unfortunately, many numerical models have no corresponding adjoint models, and developing a new one is usually a huge engineering, which consequently limits the application of the CNOP. In order to avoid adjoint models, we propose a principal components-based great deluge method to solve the CNOP. Through extracting principal components, the original problem is reduced into a low-dimensional space to hunt the coordinate of the optimal CNOP with the great deluge method. A regeneration strategy is also designed for the great deluge method to break away from local optimal positions. In addition, the proposed method can be parallelized to improve the computing efficiency. To demonstrate the validity, the proposed method has been studied in the Zebiak-Cane model to solve the CNOP. Experimental results show that the proposed method can efficiently obtain a satisfactory CNOP, approximate to the one computed with the adjoint-based method, and the parallelizing version can reach the speedup of 9.7 times with 12 cores.