Background: Clinical decision support systems (CDSS) have the potential to improve patient safety and reduce costs in primary care. However, CDSS adoption remains limited due to development and implementation challenges. CDSSs are complex interventions involving multiple interacting components that require technological innovation and behavioral and organizational change. Additionally, the primary care context is considered a complex system with high care demand, fragmented structures, and many independent yet interdependent organizations. Established determinant frameworks for implementing and scaling up complex health care interventions support the identification of implementation determinants. However, they offer limited guidance on the underlying processes of these determinants, such as the implementation processes involved in complex interorganizational collaboration in primary care. Objective: This study examined how an interorganizational collaboration in Dutch primary care ( []) achieved an iterative CDSS development and implementation. We aimed to identify the mechanisms that supported the collaboration in overcoming challenges. Methods: We performed an exploratory process-level case study. Data were collected through 15 semistructured interviews. The nonadoption, abandonment, scale-up, spread, and sustainability framework was used to ensure comprehensive topic coverage during the interviews, but not as an analytical framework. We triangulated the interviews with internal and external documents and expert input. Using a thematic, inductive approach, we developed a chronological overview of the collaboration and identified mechanisms offering insights into how GzGr navigated complexity in the development and implementation of CDSS. Results: We identified two mechanisms: (1) enacting an interorganizational value model and (2) iterative, co-creative experimentation. First, GzGr was driven by a coalition of the willing (ie, individuals willing to take an extra step), with shared goals that prioritized collective benefit while respecting organizational values. They established shared principles that translated the broad GzGr mission into concrete CDSS development choices, while also guiding strategic expansion by involving mission-aligned, innovative organizations. Second, after initial prototypes, GzGr established an iterative learning and improvement experimentation for both the technology and the collaboration. This process allowed for rapid feedback, validation of added value, and ongoing refinement. Additionally, this experimentation approached the development and implementation phase as a continuous process involving multistakeholders, supporting both the technology and the collaboration. Conclusions: This study identified 2 mechanisms that sustained interorganizational collaboration and CDSS development. These mechanisms connected collaborative and technical changes across people, technology, and organizational levels, enabling technological viability, stakeholder value, and multilevel support. The mechanisms operated both within and between organizations through iterative cycles of development and implementation. Practical implications include involving multilevel, innovative, and influential stakeholders; maintaining alignment through an orchestrating actor; and adopting an iterative approach between development and implementation. Our findings extend existing determinant frameworks by offering process-level insights into how such mechanisms help overcome challenges in the development and implementation of CDSS within interorganizational collaborations.

arXiv:2510.18212v3 Announce Type: replace Abstract: The lack of a concrete definition for Artificial General Intelligence (AGI) obscures the gap between today's specialized AI and human-level cognition. This paper introduces a quantifiable framework to address this, defining AGI as matching the cognitive versatility and proficiency of a well-educated adult. To operationalize this, we ground our methodology in Cattell-Horn-Carroll theory, the most empirically validated model of human cognition. The framework dissects general intelligence into ten core cognitive domains-including reasoning, memory, and perception-and adapts established human psychometric batteries to evaluate AI systems. Application of this framework reveals a highly "jagged" cognitive profile in contemporary models. While proficient in knowledge-intensive domains, current AI systems have critical deficits in foundational cognitive machinery, particularly long-term memory storage. The resulting AGI scores (e.g., GPT-4 at 27%, GPT-5 at 57%) concretely quantify both rapid progress and the substantial gap remaining before AGI.

arXiv:2412.05447v3 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) is one of the leading and most widely used techniques for enhancing LLM retrieval capabilities, but it still faces significant limitations in commercial use cases. RAG primarily relies on the query-chunk text-to-text similarity in the embedding space for retrieval and can fail to capture deeper semantic relationships across chunks, is highly sensitive to chunking strategies, and is prone to hallucinations. To address these challenges, we propose TOBUGraph, a graph-based retrieval framework that first constructs the knowledge graph from unstructured data dynamically and automatically. Using LLMs, TOBUGraph extracts structured knowledge and diverse relationships among data, going beyond RAG's text-to-text similarity. Retrieval is achieved through graph traversal, leveraging the extracted relationships and structures to enhance retrieval accuracy, eliminating the need for chunking configurations while reducing hallucination. We demonstrate TOBUGraph's effectiveness in TOBU, a real-world application in production for personal memory organization and retrieval. Our evaluation using real user data demonstrates that TOBUGraph outperforms multiple RAG implementations in both precision and recall, significantly improving user experience through improved retrieval accuracy.

arXiv:2510.27629v1 Announce Type: cross Abstract: Open-weight bio-foundation models present a dual-use dilemma. While holding great promise for accelerating scientific research and drug development, they could also enable bad actors to develop more deadly bioweapons. To mitigate the risk posed by these models, current approaches focus on filtering biohazardous data during pre-training. However, the effectiveness of such an approach remains unclear, particularly against determined actors who might fine-tune these models for malicious use. To address this gap, we propose \eval, a framework to evaluate the robustness of procedures that are intended to reduce the dual-use capabilities of bio-foundation models. \eval assesses models' virus understanding through three lenses, including sequence modeling, mutational effects prediction, and virulence prediction. Our results show that current filtering practices may not be particularly effective: Excluded knowledge can be rapidly recovered in some cases via fine-tuning, and exhibits broader generalizability in sequence modeling. Furthermore, dual-use signals may already reside in the pretrained representations, and can be elicited via simple linear probing. These findings highlight the challenges of data filtering as a standalone procedure, underscoring the need for further research into robust safety and security strategies for open-weight bio-foundation models.

arXiv:2510.21389v1 Announce Type: cross Abstract: Artificial intelligence (AI) systems increasingly match or surpass human experts in biomedical signal interpretation. However, their effective integration into clinical practice requires more than high predictive accuracy. Clinicians must discern \textit{when} and \textit{why} to trust algorithmic recommendations. This work presents an application-grounded user study with eight professional sleep medicine practitioners, who score nocturnal arousal events in polysomnographic data under three conditions: (i) manual scoring, (ii) black-box (BB) AI assistance, and (iii) transparent white-box (WB) AI assistance. Assistance is provided either from the \textit{start} of scoring or as a post-hoc quality-control (\textit{QC}) review. We systematically evaluate how the type and timing of assistance influence event-level and clinically most relevant count-based performance, time requirements, and user experience. When evaluated against the clinical standard used to train the AI, both AI and human-AI teams significantly outperform unaided experts, with collaboration also reducing inter-rater variability. Notably, transparent AI assistance applied as a targeted QC step yields median event-level performance improvements of approximately 30\% over black-box assistance, and QC timing further enhances count-based outcomes. While WB and QC approaches increase the time required for scoring, start-time assistance is faster and preferred by most participants. Participants overwhelmingly favor transparency, with seven out of eight expressing willingness to adopt the system with minor or no modifications. In summary, strategically timed transparent AI assistance effectively balances accuracy and clinical efficiency, providing a promising pathway toward trustworthy AI integration and user acceptance in clinical workflows.

arXiv:2510.18212v2 Announce Type: replace Abstract: The lack of a concrete definition for Artificial General Intelligence (AGI) obscures the gap between today's specialized AI and human-level cognition. This paper introduces a quantifiable framework to address this, defining AGI as matching the cognitive versatility and proficiency of a well-educated adult. To operationalize this, we ground our methodology in Cattell-Horn-Carroll theory, the most empirically validated model of human cognition. The framework dissects general intelligence into ten core cognitive domains-including reasoning, memory, and perception-and adapts established human psychometric batteries to evaluate AI systems. Application of this framework reveals a highly "jagged" cognitive profile in contemporary models. While proficient in knowledge-intensive domains, current AI systems have critical deficits in foundational cognitive machinery, particularly long-term memory storage. The resulting AGI scores (e.g., GPT-4 at 27%, GPT-5 at 57%) concretely quantify both rapid progress and the substantial gap remaining before AGI.

Cell Death Discovery, Published online: 02 September 2022; doi:10.1038/s41420-022-01165-4

Pharmacological inhibition and reversal of pancreatic acinar ductal metaplasia