Narrative comprehension on long stories and novels has been a challenging domain attributed to their intricate plotlines and entangled, often evolving relations among characters and entities. Given the LLM's diminished reasoning over extended context and its high computational cost, retrieval-based approaches remain a pivotal role in practice. However, traditional RAG methods could fall short due to their stateless, single-step retrieval process, which often overlooks the dynamic nature of capturing interconnected relations within long-range context. In this work, we propose ComoRAG, holding the principle that narrative reasoning is not a one-shot process, but a dynamic, evolving interplay between new evidence acquisition and past knowledge consolidation, analogous to human cognition on reasoning with memory-related signals in the brain. Specifically, when encountering a reasoning impasse, ComoRAG undergoes iterative reasoning cycles while interacting with a dynamic memory workspace. In each cycle, it generates probing queries to devise new exploratory paths, then integrates the retrieved evidence of new aspects into a global memory pool, thereby supporting the emergence of a coherent context for the query resolution. Across four challenging long-context narrative benchmarks (200K+ tokens), ComoRAG outperforms strong RAG baselines with consistent relative gains up to 11% compared to the strongest baseline. Further analysis reveals that ComoRAG is particularly advantageous for complex queries requiring global comprehension, offering a principled, cognitively motivated paradigm for retrieval-based stateful reasoning.

In video search systems, user historical behaviors provide rich context for identifying search intent and resolving ambiguity. However, traditional methods utilizing implicit history features often suffer from signal dilution and delayed feedback. To address these challenges, we propose WeWrite, a novel Personalized Demand-aware Query Rewriting framework. Specifically, WeWrite tackles three key challenges: (1) When to Write: An automated posterior-based mining strategy extracts high-quality samples from user logs, identifying scenarios where personalization is strictly necessary; (2) How to Write: A hybrid training paradigm combines Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO) to align the LLM's output style with the retrieval system; (3) Deployment: A parallel "Fake Recall" architecture ensures low latency. Online A/B testing on a large-scale video platform demonstrates that WeWrite improves the Click-Through Video Volume (VV$>$10s) by 1.07% and reduces the Query Reformulation Rate by 2.97%.

The automated comprehension of complex, multi-modal documents is fundamentally hampered by a disconnect between information extraction and reasoning. Existing systems suffer from inherent limitations. Monolithic models embed reasoning as a black box process, sacrificing transparency and depth. Meanwhile, current agent-based frameworks follow a passive, non-interactive paradigm; they handle static, global inputs rather than information derived from active exploration, which fundamentally restricts their ability to achieve structural understanding and complex reasoning. To bridge this critical gap, we introduce HEAR, a framework for Holistic Extraction and Agentic Reasoning. This innovative framework establishes a synergistic, closed-loop between a deep Vision-Language Model (VLM) driven holistic parsing engine and a collaborative multi-agent reasoning system. Our HEAR initially transforms unstructured documents into a semantically-rich, structured representation, preserving complex layouts and reconstituting multi-page tables. Subsequently, a multi-agent system performs cross-modal analysis, governed by a crucial verification protocol that forces agents to validate findings across textual and visual modalities. A conflict driven re-evaluation mechanism enables the system to dynamically re-engage the document to resolve ambiguities, thereby unifying the perception-cognition cycle. HEAR achieved first place in the ACM MM 2025 Grand Challenge on Large Vision–Language Model Learning and Applications.

Given the significant advances in Large Vision Language Models (LVLMs) in reasoning and visual understanding, mobile agents are rapidly emerging to meet users' automation needs. However, existing evaluation benchmarks are disconnected from the real world and fail to adequately address the diverse and complex requirements of users. From our extensive collection of user questionnaire, we identified five tasks: Multi-App, Vague, Interactive, Single-App, and Unethical Instructions. Around these tasks, we present MVISU-Bench, a bilingual benchmark that includes 404 tasks across 137 mobile applications. Furthermore, we propose Aider, a plug-and-play module that acts as a dynamic prompt prompter to mitigate risks and clarify user intent for mobile agents. Our Aider is easy to integrate into several frameworks and has successfully improved overall success rates by 19.55% compared to the current state-of-the-art (SOTA) on MVISU-Bench. Specifically, it achieves success rate improvements of 53.52% and 29.41% for unethical and interactive instructions, respectively. Through extensive experiments and analysis, we highlight the gap between existing mobile agents and real-world user expectations.