Qin Jin

This paper explores attention attractors —tokens that draw significantly high attention—in large language models. We analyze them from three perspectives: (1) Functionality : We demonstrate their role in aggregating information from preceding contexts to facilitate future predictions. (2) Distrib...

We present HowToNarrate , the first general-domain benchmark for Synchronized Video Narration. The benchmark contains 3.2K videos across seven domains, segmented into 37.5K clips with aligned narrations and associated external knowledge. Effective narration requires models to understand visual sc...

Geometry problem solving, a crucial aspect of mathematical reasoning, is vital across various domains, including education, the assessment of AI’s mathematical abilities, and multimodal capability evaluation. The recent surge in deep learning technologies, particularly the emergence of multimodal...

Charts are a universally adopted medium for data communication, yet existing chart understanding benchmarks are overwhelmingly English-centric, limiting their accessibility and relevance to global audiences. To address this limitation, we introduce PolyChartQA, the first large-scale multilingual...

Understanding multimodal user personas in long-term dialogues is essential for building personalized and human-like dialogue systems. However, existing datasets suffer from limited persona diversity and static, overly simplified settings, making them insufficient for capturing the complexity of r...

Sports have witnessed growing global enthusiasm in recent years, serving as a vital force for physical health, cultural exchange, social connection, and economic growth. The rapid advancement of large models, particularly (multimodal) large language models (M)LLMs, has demonstrated transformative...

Automatic evaluation of book-length stories remains underexplored, particularly for non-English literature. We introduce ChangJuan, the first benchmark for book-length Chinese story evaluation , comprising 300 novels with metadata, human ratings, and large-scale user reviews. To mitigate the subj...