Abstract: The Third Plenary Session of the Twentieth Central Committee of the Communist Party of China proposed “improving the institutional mechanisms for developing new quality productive forces in accordance with local conditions,” providing a strategic guideline for advancing the transformation and upgrading of regional productivity. As a new form of productive force characterized by innovation-driven dynamics and efficiency-oriented goals, the development of new quality productive forces is crucial for enhancing regional vitality and restructuring economic systems toward high-quality development. However, the implementation of this national strategy at the local level has revealed significant challenges, particularly the trend of policy convergence and homogeneous industrial planning across regions, which risks undermining local comparative advantages and aggravating inefficient competition. Drawing on the theory of government attention, which posits that the allocation of attention determines the prioritization of policy resources and institutional efforts, this study investigates how local governments interpret and implement the central directive on new quality productive forces. Specifically, the study collects and analyzes political affairs texts related to new quality productive forces, published on the official websites of thirty provincial-level governments between 2023 and 2024. Using topic modeling based on the BERTopic algorithm, generative artificial intelligence techniques, and machine learning methods such as the random forest algorithm, the study systematically identifies the focus areas of government attention and examines their regional variations. The findings reveal three major patterns. First, local governments demonstrate a structure of “five-dimensional focus and coexistence of dual tendencies” in their attention, covering five domains: new means of labor, new types of laborers, industrial sectors, application scenarios, and enabling environments. This indicates both differentiated governance strategies and converging policy orientations across regions. Second, policy imitation behavior is prevalent and primarily manifests as “late-developing region catch-up” and “peer-level learning.” This results in a spatial paradox in which industrial policy similarity increases with geographical distance, reflecting a widespread tendency to emulate economically advanced regions regardless of local resource endowments or development conditions. Third, regional heterogeneity in governance strategies is evident. Eastern provinces, which possess strong innovation capabilities and relatively high fiscal autonomy, tend to demonstrate desirable convergence characterized by targeted adaptation. In contrast, central provinces, constrained by weak industrial foundations and fragmented development trajectories, are more prone to undesirable convergence and may fall into a governance dilemma of “resource involution.” To address the risks of excessive policy convergence and foster truly localized development of new quality productive forces, this study proposes three governance recommendations. First, the central government should strengthen top-level design, promote layered and classified policy guidance, improve regional coordination mechanisms, and align national strategies with the specific conditions and needs of localities. Second, local governments should enhance regional cooperation and policy synergy, avoid blind imitation, and build cross-regional industrial cooperation platforms to promote rational division of labor and resource allocation. Third, governments at all levels should coordinate differentiated innovation strategies in key industries with the collaborative development of related sectors, clarify industrial roles based on regional development stages, and prevent inefficient duplication. These recommendations aim to support a more adaptive, efficient, and coordinated institutional framework for the development of new quality productive forces in China.