Background: Online consultation platforms have become an important component of survivorship care for patients with cancer, offering flexible access to oncology expertise between scheduled visits. However, evidence on what drives the willingness of survivors of cancer to continue using online consultations after initial adoption remains limited in China. A better understanding of continuance intention is needed to inform survivor-centered digital health strategies. Objective: This study aimed to explore the influencing factors of continued use of online consultations among survivors of cancer in southwest China and develop a grounded theoretical model explaining continuance intention. Methods: A grounded theory qualitative design was used. A total of 26 adult survivors of cancer with diverse demographic and clinical characteristics were purposively recruited from a tertiary cancer center in southwest China. All participants had used online consultations at least once in the preceding year. Semistructured telephone interviews were audio recorded; transcribed verbatim; and analyzed using open, axial, and selective coding with constant comparison until theoretical saturation was reached. During selective coding, categories and their relationships were integrated and iteratively refined to construct a grounded theoretical model of continuance intention. Results: Six interrelated domains influenced survivors’ continued use of online consultation platforms: platform quality, physician competence, user perception, individual condition, external context, and privacy concerns. Platform quality and physician competence influenced user perception of usefulness, reassurance, and trust, which functioned as a mediator of continued use. Individual condition, including health status, health literacy, and psychological needs, influenced both perceived usefulness and reliance on online consultations. External context, especially family encouragement, peer recommendations, and availability of local oncology services, directly facilitated or constrained continued use. Privacy concerns moderated how survivors balanced perceived benefits against risks of data misuse, stigma, and unwanted disclosure of cancer history. Survivors described online consultations as offering rapid guidance and emotional support that complemented hospital-based care but reported discontinuation when interactions were delayed or impersonal or when perceived privacy risks outweighed the benefits. Conclusions: The willingness of survivors of cancer to continue using online consultation platforms depends on multiple interrelated factors beyond traditional technological usability. Sustained engagement is shaped by survivors’ perceptions of usefulness and trust, physician empathy and timeliness, family encouragement, and acceptance of privacy trade-offs. The theoretical model advances understanding of digital health continuance in oncology and offers practical guidance for developing survivor-centered online consultation services.

arXiv:2511.05385v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG) utilizes external knowledge to augment Large Language Models' (LLMs) reliability. For flexibility, agentic RAG employs autonomous, multi-round retrieval and reasoning to resolve queries. Although recent agentic RAG has improved via reinforcement learning, they often incur substantial token overhead from search and reasoning processes. This trade-off prioritizes accuracy over efficiency. To address this issue, this work proposes TeaRAG, a token-efficient agentic RAG framework capable of compressing both retrieval content and reasoning steps. 1) First, the retrieved content is compressed by augmenting chunk-based semantic retrieval with a graph retrieval using concise triplets. A knowledge association graph is then built from semantic similarity and co-occurrence. Finally, Personalized PageRank is leveraged to highlight key knowledge within this graph, reducing the number of tokens per retrieval. 2) Besides, to reduce reasoning steps, Iterative Process-aware Direct Preference Optimization (IP-DPO) is proposed. Specifically, our reward function evaluates the knowledge sufficiency by a knowledge matching mechanism, while penalizing excessive reasoning steps. This design can produce high-quality preference-pair datasets, supporting iterative DPO to improve reasoning conciseness. Across six datasets, TeaRAG improves the average Exact Match by 4% and 2% while reducing output tokens by 61% and 59% on Llama3-8B-Instruct and Qwen2.5-14B-Instruct, respectively. Code is available at https://github.com/Applied-Machine-Learning-Lab/TeaRAG.

arXiv:2510.24701v1 Announce Type: cross Abstract: We present Tongyi DeepResearch, an agentic large language model, which is specifically designed for long-horizon, deep information-seeking research tasks. To incentivize autonomous deep research agency, Tongyi DeepResearch is developed through an end-to-end training framework that combines agentic mid-training and agentic post-training, enabling scalable reasoning and information seeking across complex tasks. We design a highly scalable data synthesis pipeline that is fully automatic, without relying on costly human annotation, and empowers all training stages. By constructing customized environments for each stage, our system enables stable and consistent interactions throughout. Tongyi DeepResearch, featuring 30.5 billion total parameters, with only 3.3 billion activated per token, achieves state-of-the-art performance across a range of agentic deep research benchmarks, including Humanity's Last Exam, BrowseComp, BrowseComp-ZH, WebWalkerQA, xbench-DeepSearch, FRAMES and xbench-DeepSearch-2510. We open-source the model, framework, and complete solutions to empower the community.

arXiv:2509.13345v2 Announce Type: replace-cross Abstract: Hallucination in generative AI is often treated as a technical failure to produce factually correct output. Yet this framing underrepresents the broader significance of hallucinated content in language models, which may appear fluent, persuasive, and contextually appropriate while conveying distortions that escape conventional accuracy checks. This paper critically examines how regulatory and evaluation frameworks have inherited a narrow view of hallucination, one that prioritises surface verifiability over deeper questions of meaning, influence, and impact. We propose a layered approach to understanding hallucination risks, encompassing epistemic instability, user misdirection, and social-scale effects. Drawing on interdisciplinary sources and examining instruments such as the EU AI Act and the GDPR, we show that current governance models struggle to address hallucination when it manifests as ambiguity, bias reinforcement, or normative convergence. Rather than improving factual precision alone, we argue for regulatory responses that account for languages generative nature, the asymmetries between system and user, and the shifting boundaries between information, persuasion, and harm.