Das L3S ist ein Forschungszentrum für grundlagen- und anwendungsorientierte Forschung im Bereich Digitale Transformation sowie Data- und Web-Science. L3S-Forscher entwickeln zukunftsweisende Methoden und Informationstechnologien für intelligente, zuverlässige und verantwortungsvolle Systeme, welche zur Gestaltung der digitalen Transformation und zur nachhaltigen Innovation beitragen. Der Wissentransfer in Wirtschaft, Politik und Gesellschaft bildet einen zentralen Baustein der Aktivitäten des L3S.

Das L3S erforscht die Auswirkungen des digitalen Wandels, um aus den Erkenntnissen Handlungsoptionen, -empfehlungen sowie Innovationsstrategien für die Wirtschaft, die Politik und Gesellschaft herzuleiten. Durch Forschung, Entwicklung und Beratung trägt das L3S gemeinsam mit seinen Partnern zur digitalen Transformation insbesondere in den Bereichen Mobilität, Gesundheit, Produktion und Bildung bei.

Das L3S ist eine gemeinsame Einrichtung der Leibniz Universität Hannover und der Technischen Universität Braunschweig mit rund 150 Wissenschaftlern aus unterschiedlichen Disziplinen an Universitäten und einem Leibniz-Institut (der TIB). Das aktuelle jährliche Drittmittelvolumen des L3S beträgt etwa 6 Mio. Euro, Niedersachsen fördert das L3S zusätzlich mit jährlich etwa 2 Mio. Euro. Das L3S wurde 2011 und 2014 vom Wissenschaftsrat als exzellente Einrichtung evaluiert, führend im Bereich Web Science.

• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• Concrete field margins and insecticide use erode and restructure spider assemblages in subtropical smallholder landscapes. Long, Jing; Boetzl, Fabian A.; Zou, Yi; Chen, Yanjie; Zhang, Jie; Li, Yiming; Stör, Jean-Léonard; Zhang, Fusuo; Zhang, Chaochun; Steffan-Dewenter, Ingolf (2027). 414 110788.
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• Concrete field margins and insecticide use erode and restructure spider assemblages in subtropical smallholder landscapes. Long, Jing; Boetzl, Fabian A.; Zou, Yi; Chen, Yanjie; Zhang, Jie; Li, Yiming; Stör, Jean-Léonard; Zhang, Fusuo; Zhang, Chaochun; Steffan-Dewenter, Ingolf (2027). 414 110788.
• MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks. Borst, Vanessa; Horn, Lukas; Grillmeyer, Daniel; Prantl, Thomas; Kounev, Samuel (2027).
• Teleoperation and Multi-Hop Communication Architecture of a Mars Cave Rover via Satellite Interface: Results from Terrestrial Analog Tests. Maurer, Andreas; Koch, Johannes Frithjof; Ruschinzik, Lukas; Schütt, Manuel (2026).
• Structured Gaussian Processes for Uncertainty-Aware Classification of High-Dimensional, Small-Sampled Omics Data. Zhang, Yue; Gadhia, Nandini Amit; Karagiannis, Georgios; Smyrnakis, Michalis P. Veltri, G. Pozzi, P. Vizza, M. Mantovani, P. H. Guzzi, Y. Luo, L. Shen (eds.) (2026). 10:1–10:10.
• Sustainability as a Part of QoE? Crowdsourced Insights on Streaming Quality Compromises. Baur, Carina; Schaaf, Moritz; Böttger, Friedrich; Wehner, Nikolas; Poignée, Fabian; Loh, Frank; Hoßfeld, Tobias (2026).
• CASSANDRA hört zu und hilft: Ein KI-Teammitglied für die Anästhesiologie. Huber, Stephan; Grundgeiger, Tobias (2026).
• From Nuisance to Facilitator: A Mixed-Methods Evaluation of an AI Documentation Assistant for Anesthesiologists in the Operating Room. Grundgeiger, Tobias; Huber, Stephan; Wießner, Adriana; Dunkelberg, Matthias; Ottenhaus, Simon; Nagel, Anja; Dilling, Christina; Happel, Oliver (2026).
• Synaptic connectome of a neurosecretory network in the textitDrosophila brain. McKim, Theresa H; Gera, Jayati; Gayban, Ariana J; Christie, Kevin; Shin, Joel; Reinhard, Nils; Manoli, Giulia; Hilpert, Selina; Callaerts, Patrick; van Breugel, Floris; Helfrich-Förster, Charlotte; Zandawala, Meet (2026). 13 RP102684.
• Prosumers at the Crossroads of Policy, Practice, and Public Discourse: An Exploratory Analysis of the Romanian Context. Zbuchea, Alexandra; Barna, Cristina; Stănescu, Simona; Rebeleanu, Adina B. Hrib, A. Mitan, P. Mititean, L. Oprea, A.-M. Osadci-Baciu, A. Zbuchea (eds.) (2026). 391–410.

  The paper explores the phenomenon of prosumers within the Romanian context. Positioned at the intersection of policy, practice, and public discourse, the study examines how national regulatory frameworks, market dynamics, and media narratives collectively shape the development of prosumerism as a significant component of the energy transition. Methodologically, the research employs a mixed-methods exploratory design, combining qualitative policy and discourse analysis with quantitative analysis of sectoral data provided by the National Energy Regulatory Authority and the Romanian National Institute of Statistics. The analysis covers the period 2019–2025, highlighting key momentum in the evolution of prosumers and energy communities. Main findings reveal a rapid expansion of the prosumer sector, driven primarily by two support mechanisms: the Casa Verde Fotovoltaice investment program and quantitative compensation. Despite the expansion, significant challenges persist, including regulatory fragmentation, infrastructural constraints, and regional disparities in adoption. The emergence of the Association of Prosumers and Energy Communities in 2024 marks an important institutional step toward collective representation and policy advocacy. Media analysis indicates that public discourse around prosumers has shifted from marginal awareness to a broader framing of energy citizenship, autonomy, and participation. However, some tensions remain between policy rhetoric and implementation realities. Overall, the study concludes that prosumers have begun to become central stakeholders in Romania’s energy transition. Also, by integrating multiple perspectives, our study contributes to a comprehensive understanding of how prosumerism can accelerate the transition towards a prosumer ecosystem that better aligns with EU objectives for a decentralized, just, and citizen-driven energy transition.
• A Causality-Inspired Spatial-Temporal Return Decomposition Approach for Multi-Agent Reinforcement Learning. Zhang, Yudi; Du, Yali; Huang, Biwei; Fang, Meng; Pechenizkiy, Mykola V. Mascardi, J. Thangarajah, C. Amato, L. A. Dennis (eds.) (2026). 3238–3240.
• HUAL-PAF: Boosting Multimodal Sentiment Analysis with Harmless Unimodal Auxiliary Learning and Progressive Attention Fusion. Li, Jing; Zhang, Xuejie; Wang, Jin; Zhou, Xiaobing in Lecture Notes in Computer Science, D.-S. Huang, C. Zhang, W. Chen, Y. Pan, P. Premaratne (eds.) (2026). (Vol. 16665) 54–65.
• Generative Bayesian Optimization: Generative Models as Acquisition Functions. Oliveira, Rafael; Steinberg, Daniel M.; Bonilla, Edwin V. C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• In-Context Compositional Q-Learning for Offline Reinforcement Learning. Xu, Qiushui; Huang, Yuhao; Jiang, Yushu; Zheng, Wenliang; Song, Lei; Wang, Jinyu; Bian, Jiang C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• Scheduling Your LLM Reinforcement Learning with Reasoning Trees. Wang, Hong; Hao, Zhezheng; Luo, Jian; Wei, Chenxing; Shu, Yao; Liu, Lei; Lin, Qiang; Dong, Hande; Chen, Jiawei C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• Generation then Reconstruction: Accelerating Masked Autoregressive Models via Two-Stage Sampling. Yan, Feihong; Zhu, Yao; Wang, Peiru; Pang, Kaiyu; Wei, Qingyan; Li, Huiqi; Zhang, Linfeng C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• HiDrop: Hierarchical Vision Token Reduction in MLLMs via Late Injection, Concave Pyramid Pruning, and Early Exit. Wu, Hao; Fan, Yingqi; Dai, Jinyang; Tong, Junlong; Ma, Yunpu; Shen, Xiaoyu C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• PostAlign: Multimodal Grounding as a Corrective Lens for MLLMs. Wu, Yixuan; Zhang, Yang; Wu, Jian; Torr, Philip; Gu, Jindong C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer. Li, Yanghao; Qian, Rui; Pan, Bowen; Zhang, Haotian; Huang, Haoshuo; Zhang, Bowen; Tong, Jialing; You, Haoxuan; Du, Xianzhi; Gan, Zhe; Kim, Hyunjik; Jia, Chao; Wang, Zhenbang; Yang, Yinfei; Gao, Mingfei; Dou, Zi-Yi; Hu, Wenze; Gao, Chang; Li, Dongxu; Dufter, Philipp; Wang, Zirui; Yin, Guoli; Zhang, Zhengdong; Chen, Chen; Zhao, Yang; Pang, Ruoming; Chen, Zhifeng C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• Multiplayer Nash Preference Optimization. Wu, Fang; Huang, Xu; Xuan, Weihao; Zhang, Zhiwei; Xiao, Yijia; Wan, Frank; Li, Xiaomin; Hu, Bing; Xia, Peng; Leskovec, Jure; Choi, Yejin C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• Learnable Sparsity for Vision Generative Models. Zhang, Yang; Jin, Er; Liang, Wenzhong; Dong, Yanfei; Khakzar, Ashkan; Torr, Philip; Stegmaier, Johannes; Kawaguchi, Kenji C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• Long-Context Generalization with Sparse Attention. Vasylenko, Pavlo; Pitorro, Hugo; Martins, André F. T.; Treviso, Marcos V. C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• How the EU Deforestation Policy Could Reduce Forest Loss Around the World. Carter, Sarah; Li, Bo; Campbell, Olivia; Zantow, Sonja (2026).
• VaMEx3-MarsSymphony: Enabling Robust Mars Cave Exploration with Distributed Sensor Networks. Maurer, Andreas; Johne, Niklas; Junker, Julian; Kayal, Hakan; Mutter, Julian; Riegler, Clemens; Ryzhykh, Daria (2026).
• VaseVQA-3D: Benchmarking 3D VLMs on Ancient Greek Pottery. Zhang, Nonghai; Zhang, Zeyu; Wang, Jiazi; Zhao, Yang; Tang, Hao C. Vondrick, B. Hariharan, C. Raffel, L. Pinto, D. Yang, A. Faust (eds.) (2026).
• CreoBench: An Open-Loop, Distributed Load Tester for Modern Microservice Benchmarking. Lubas, Yannik; Straesser, Martin; Rohwer, Ivo; Kounev, Samuel (2026). (Vol. 46)

  Open-loop, distributed load generation is a well-established pattern for realistic microservice benchmarking, yet implementations remain scarce. Existing tools force practitioners into unsatisfying trade-offs. Single-host open-loop generators bottleneck on one machine, while distributed closed-loop frameworks suffer from coordinated omission. Existing open-loop distributed load testers introduced useful architectural patterns, but their implementations have not kept pace with modern workload demands. Building on these principles, we present CreoBench, an open-loop distributed load tester written in Rust. In this paper, we provide (i) our design decisions for distributed, open-loop load generation including precisely paced Poisson arrivals, (ii) an open-source implementation we hope the community can collectively improve, and (iii) an empirical evaluation demonstrating that our tool generates Poisson arrivals with sub-millisecond precision, and comparing its resource usage to the original Java implementation.
• Serverless Workflow Engineering for Large-Scale Interdisciplinary Data Analysis in Earth Observation. Kounev, Samuel (2026, October 5).
• Serverless Workflow Engineering for Large-Scale Interdisciplinary Data Analysis in Earth Observation. Kounev, Samuel (2026, October 5).
• Dynamics of charge-density-wave puddles in 2H-NbSe\($_\mathbf2$\). Kumbhakar, Shreya; Esposito, Marina; Kumar N M, Anjan; Feng, Liwen; Confalone, Tommaso; Gerega, Vasilisa; Lo Sardo, Flavia; Alam, Rafiqul; Masarotti, Davide; Tafuri, Francesco; Böhm, Thomas; Abdel-Hafiez, Mahmoud; Chandra, Sushmita; Felser, Claudia; Nielsch, Kornelius; Poccia, Nicola; Kaiser, Stefan; Haider, Golam (2026).

  Electronic phases in quantum materials are often governed by nanoscale inhomogeneity, where local order develops within spatially confined regions, or puddles. A prominent example is an incommensurate charge-density-wave (I-CDW) comprising locally commensurate domains. In 2H-NbSe2, such an I-CDW state persists alongside lattice anharmonicity and superconductivity, raising fundamental questions about the dynamical stabilization of CDW order in puddles. Here, using Raman scattering, we reveal a strong Fano-coupling between interlayer shear phonons and the CDW amplitude mode. Our time-resolved reflectivity measurement shows a low-frequency ∼0.2 THz coherent overdamped oscillation onsetting within the CDW regime at ∼17K, indicating a previously unexplored transition. This we identify as a Fano-coupled phonon-CDW hybrid emerging from the collective dynamics of CDW puddles. We establish an overdamped-oscillator framework for capturing such intertwined glassy dynamics. These results highlight how lattice pinning and electronic correlations shape CDW order in layered materials, insight crucial for designing novel van der Waals devices.
• An Area Efficient and Faster Sign Detector for the Moduli Set 2n, 22n−1, 22n+1. Technical Report (PhD dissertation), Bankas, Edem Kwedzo; Bilal, Yakubu (2026).
• GP-VM×SMA: Benchmarking General-Purpose Vision Models and Specialized Architectures for 2D Medical Image Segmentation. Borst, Vanessa; Riedmann, Anna; Kounev, Samuel (2026).
• SEQUENCE-TO-SEQUENCE DEEP LEARNING FOR DEPENDENCY-AWARE TASK OFFLOADING IN MULTI-ACCESS EDGE COMPUTING. Elhajji, Mohamed; Chaou, Abdessamad; Lagnfdi, Oussama; Darif, Anouar (2026). 18(5) 37–50.

  The proliferation of Internet of Things (IoT) devices and resource-intensive applications has placed growing pressure on computation, storage, and network resources. Cloud computing helps, but its centralized nature adds latency that delay-sensitive services cannot tolerate, whereas Multi-Access Edge Computing (MEC), boosted by 5G, moves computation closer to the data. Offloading remains hard, however, because applications consist of interdependent tasks while wireless conditions and edge load fluctuate continuously. This paper formulates dependency-aware task offloading as a Sequence-toSequence (Seq2Seq) learning problem: the application task graph is read as an input sequence, and pertask placements are produced as an output sequence. The contributions are threefold: (i) a Seq2Seq encoder-decoder model that couples a graph-aware encoder with multi-scale (global, local, and criticalpath) attention; (ii) an offline curriculum-based training procedure that removes the online exploration cost of reinforcement learning; and (iii) an extensive evaluation on a discrete-event MEC simulator. The proposed model reduces application latency by up to 38% compared with Round-Robin and by about 26% compared with HEFT, with the widest margin in the 7–20 Mbps band typical of 4G networks; it scales near-linearly at about 1.7 ms per task up to 100 tasks; and it keeps the task-failure rate near 5% under overload, against 12% for Round-Robin, at a modest 18% device-energy overhead.
• Real-Time Ground Normal Vector Estimation for Spherical Mobile 3D Mapping using LiDAR-Inertial Odometry. Schubert, T.; Arzberger, F.; Nüchter, A. (2026).

  Accurate ground plane normal estimation is essential for odometry in spherical mobile mapping systems (SMMS), which face challenges from rapid rotations and limited ground coverage. We propose two RANSAC-based approaches: single-scan estimation with normal vector scoring and multi-step validation, and LiDAR-inertial odometry (LIO)-based estimation using a global 3D map. Evaluated across diverse real-world datasets against terrestrial laser scanner ground truth, results show that LIO- based estimation significantly outperforms single-scan methods in accuracy and robustness. Integrating ground normal estimates into the motion model improves pose estimation, supporting autonomous navigation in challenging and hazardous environments.
• Serverless Workflow Engineering for Large-Scale Interdisciplinary Data Analysis in Earth Observation. Kounev, Samuel (2026, October 5).
• Do Pedagogical Agent Roles Matter? Effects on Learning Performance, Motivation, and Agent Perception. Riedmann, Anna; Lugrin, Birgit in IVA 2026 (2026).

  Pedagogical Agents (PAs) have been widely studied in technology-supported learning environments and are generally considered beneficial for enhancing students’ learning and motivation. However, less is known about how specific agent roles influence these outcomes. This work addresses this gap by examining the effects of different PA roles on learning performance, motivation, and learners’ perception of the PA. In a user study, two role variants — a mentor and an expert agent — were compared, alongside a control condition without a PA. The results showed no significant differences in learning performance or motivation between conditions. However, the mentor agent was perceived as significantly more facilitative of learning, more human-like, and more engaging than the expert agent. These findings suggest that agent role design primarily seems to shape learners’ subjective experience, without necessarily leading to measurable improvements in learning outcomes. Overall, the results do not confirm the beneficial effects of PAs on learning outcomes and motivation compared to a non-agent control condition, contrary to what is reported in prior research. However, when a PA is implemented, the mentor role appears to be the more suitable design choice, as it enhances the affective and social aspects of the learning experience.
• Well-ordering principles over recursive comprehension. Buriola, Gabriele; Cozzi, Giacomo; Dammrau, Antonia; Freund, Anton (2026). 27 pp.
• Social Robots in Higher Education: Introducing an Adaptive Robotic Tutor for Long-term Interaction Donnermann, Melissa PhD thesis, [Doctoral Thesis]. Universität Würzburg. (2026).
• In-orbit monitoring for dynamic detector-oriented containment of TID-induced degradation in CMOS-TDI sensor systems. https://doi.org/10.1038/s41598 026-72300-9 (2026).

  Space-borne CMOS-based sensor systems with time delay and integration (TDI) functionality are susceptible to the radiation environment in orbit. Total ionizing dose (TID) impacts CMOS detector structures and considerably deteriorates performance and image quality. We investigated dynamic detector-oriented methods to counteract TID-induced degradation effects. We equipped the CMOS-TDI detector system with in-orbit monitoring capabilities applying charge injection and subsampling to reveal inherent detector operation feedback and enable system adaptation during operation. Thus, the impact of radiation was significantly contained. The charge transfer efficiency (CTE) accordingly reached the pre-radiation level. The relative charge loss was reduced by a factor of 8.1, yielding a substantial improvement in CTE, and hence, in the geometric resolution. We further applied device compact modeling, aiming for a comprehensive system model to facilitate adequate prediction of performance degradation caused by TID, which could be generally applied to upcoming designs of the CMOS-TDI detector family after external validation.
• Ballistic electron beams in \($\mathrmPdCoO_2$\). Aravindnath, Abu Alex; Pyshkin, Pavlo V.; König, Markus; Khim, Seunghyun; Kleinlein, Johannes; Buhmann, Hartmut; Molenkamp, Laurens W. (2026). 114(23) L231404.

  The transport properties of \($\mathrmPdCoO_2$\), a material with a hexagon-shaped Fermi surface, are investigated in the two contrasting crystal directions by ballistic electron deflection in a perpendicular magnetic field. By progressively reducing the device size, we systematically examine the impact of geometric confinement on the electron propagation. The measurements reveal clear orientation-dependent features. The quantitative and qualitative agreement with a theoretical model based on the Landauer-B"uttiker formalism allows one to distinguish features originating from classical size effects and from the intrinsic anisotropy of the Fermi surface.
• October 2026: Top 10 Read Articles in Computer Networks & Communications. IJCNC (2026).

  The International Journal of Computer Networks & Communications (IJCNC) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of Computer Networks & Communications. The journal focuses on all technical and practical aspects of Computer Networks & data Communications. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced networking concepts and establishing new collaborations in these areas. Authors are solicited to contribute to this journal by submitting articles that illustrate research results, projects, surveying works and industrial experiences that describe significant advances in the Computer Networks & Communications.
• Culture-Specific Verbal Feedback by Robot Tutors: A Cross-Cultural Study in Rwanda and Germany. Donnermann, Melissa; Vernon, David; Lugrin, Birgit (2026). 1315–1322.
• October 2026 - Top Cited Articles of International Journal on Soft Computing, Artificial Intelligence and Applications (IJSCAI). IJSCAI (IJSCAI, ed.) (2026).

  International Journal on Soft Computing, Artificial Intelligence and Applications (IJSCAI) is an open access peer-reviewed journal that provides an excellent international forum for sharing knowledge and results in theory, methodology and applications of Artificial Intelligence, Soft Computing. The Journal looks for significant contributions to all major fields of the Artificial Intelligence, Soft Computing in theoretical and practical aspects. The aim of the Journal is to provide a platform to the researchers and practitioners from both academia as well as industry to meet and share cutting-edge development in the field.
• Ethical Guidelines and Stakeholder Perspectives on the Use of Conversational AI in Psychotherapy - A Scoping Review. Friedrich, Paula; Obremski, David; Latoschik, Marc Erich; Wienrich, Carolin in ICMI ’26 (2026). 15–26.

  The global burden of mental health disorders has created considerable demand for mental health professionals and long waiting periods for affected individuals. Due to advancements in generative Artificial Intelligence (AI) in recent years, it is now considered capable of simulating meaningful social interaction. Hence, a growing amount of research has investigated the use of conversational AI in the area of mental health to complement the work of therapists and increase access to care. To mitigate the risks that come with using AI in psychotherapy, a broad variety of ethical considerations from different fields have been introduced; however, stakeholder perspectives regarding this development still remain under-researched. Therefore, we first break down broad theoretical and empirical research to facilitate ethical design for developers and identify six overarching ethical principles: Beneficence, Risk and Harm Avoidance, Data Autonomy, Fairness, Involvement, and Responsibility. Second, we review 74 papers (2018-2025) discussing the evaluation of stakeholder perspectives in the current literature. Lastly, we merge our findings and provide stakeholder-focused implications for the human-centered ethical development of AI in psychotherapy.
• Enabling Unstructured Sparse Fine-Tuning and Inference for Foundation Models on Wafer-Scale Engine. Zheng, Haoyu; Zeng, Yifan; Song, Linghao; Emani, Murali; Dong, Wenqian (2025). 500–507.
• Modeling Higher-order Human Beliefs Using the Justified Perspective Model. Li, Wanchun; Zhang, Chenyuan; Li, Weijia; Hu, Guang; Xu, Yangmengfei N. Yamashita, V. Evers, K. Yatani, S. X. Ding (eds.) (2025). 404:1–404:8.
• Coft: Making Large Language Models Better Zero-Shot Learners for Code Generation. Li, Weijia; Qian, Yongjie; Gao, Ke; Chen, Haixin; Wang, Xinyu; Tong, Yuchen; Li, Ling; Wu, Yanjun; Zhao, Chen (2025). 489–499.
• Stop Looking for "Important Tokens" in Multimodal Language Models: Duplication Matters More. Wen, Zichen; Gao, Yifeng; Wang, Shaobo; Zhang, Junyuan; Zhang, Qintong; Li, Weijia; He, Conghui; Zhang, Linfeng C. Christodoulopoulos, T. Chakraborty, C. Rose, V. Peng (eds.) (2025). 9961–9980.
• RuAG: Learned-rule-augmented Generation for Large Language Models. Zhang, Yudi; Xiao, Pei; Wang, Lu; Zhang, Chaoyun; Fang, Meng; Du, Yali; Puzyrev, Yevgeniy; Yao, Randolph; Qin, Si; Lin, Qingwei; Pechenizkiy, Mykola; Zhang, Dongmei; Rajmohan, Saravan; Zhang, Qi (2025).
• Umkämpfte Transformation: Konflikte um den digitalen und ökologischen Wandel Dörre, Klaus; Liebig, Steffen; Lucht, Kim; Michaelis, Lennart; Sittel, Johanna in International Labour Studies (2025). Campus-Verlag, Weinheim.

  Der ökologische und digitale Wandel der Gesellschaft befindet sich gegenwärtig im vollen Gange – und ist doch umkämpft wie nie zuvor. Immer deutlicher treten manifeste Konflikte um die Verteilung der Kosten und Lasten der notwendigen Transformation auf den Plan, während gleichzeitig um die Richtung und Tiefe des Wandels gerungen wird. Mitunter drohen soziale Interessen gegen ökologische Anliegen ausgespielt zu werden – und vice versa. Zuletzt führte die unzureichende Berücksichtigung sozialer Anliegen verstärkt zu anti-ökologischen Haltungen und bedrohlichen Erfolgen rechter Parteien. Die Beiträge des Sammelbandes, welche überwiegend aus der Forschung am Arbeitsbereich für Arbeits-, Industrie- und Wirtschaftssoziologie der Friedrich-Schiller-Universität Jena hervorgehen, thematisieren diese Transformationskonflikte anhand unterschiedlicher Fallstudien. Ausgehend von konzeptuellen Beiträgen zu Transformation und Klasse, wird der umkämpfte ökologische und digitale Wandel in ausgewählten Branchen empirisch untersucht. Der Band schließt mit Vorschlägen zur demokratischen Bearbeitung der analysierten Konflikte aus (industrie-)politischer, gewerkschaftlicher und ökologischer Perspektive.
• VisualEDU: A Benchmark for Assessing Coding and Visual Comprehension through Educational Problem-Solving Video Generation. Chen, Hao; Shi, Tianyu; Huang, Pengran; Li, Zeyuan; Pan, Jiahui; Chen, Qianglong; He, Lewei C. Christodoulopoulos, T. Chakraborty, C. Rose, V. Peng (eds.) (2025). 16363–16394.
• Spiral of Silence in Large Language Model Agents. Zhong, Mingze; Fang, Meng; Shi, Zijing; Huang, Yuxuan; Zheng, Shunfeng; Du, Yali; Chen, Ling; Wang, Jun C. Christodoulopoulos, T. Chakraborty, C. Rose, V. Peng (eds.) (2025). 23238–23253.
• In2X at WMT25 Translation Task. Pang, Lei; Mao, Hanyi; Xiao, Quanjia; Chen, Ruihan; Zhang, Jingjun; Liu, HaiXiao; Li, Xiangyi B. Haddow, T. Kocmi, P. Koehn, C. Monz (eds.) (2025). 671–679.
• Effects of Information Widgets on Time Perception during Mentally Demanding Tasks. Li, Zengrui; Shi, Di; Gao, Qijun; Chen, Yichen; Wang, Nanyi; Ren, Xipei N. Yamashita, V. Evers, K. Yatani, S. X. Ding, B. Lee, M. Chetty, P. O. T. Dugas (eds.) (2025). 753:1–753:20.
• Risk-Aware Planner for Quadrotor in Cluttered and Dynamic Environments. Li, Yongjian; Zheng, Minzhe; Chen, Kai; Liu, Hongji; Zhou, Jinni; Wang, Lujia; Ma, Jun (2025). 1009–1015.
• VIGC: Visual Instruction Generation and Correction. Wang, Bin; Wu, Fan; Han, Xiao; Peng, Jiahui; Zhong, Huaping; Zhang, Pan; Dong, Xiaoyi; Li, Weijia; Li, Wei; Wang, Jiaqi; He, Conghui M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 5309–5317.
• Multi-Prototype Space Learning for Commonsense-Based Scene Graph Generation. Chen, Lianggangxu; Song, Youqi; Cai, Yiqing; Lu, Jiale; Li, Yang; Xie, Yuan; Wang, Changbo; He, Gaoqi M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 1129–1137.
• S2CycleDiff: Spatial-Spectral-Bilateral Cycle-Diffusion Framework for Hyperspectral Image Super-resolution. Qu, Jiahui; He, Jie; Dong, Wenqian; Zhao, Jingyu M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 4623–4631.
• Analysis of the Impacts of Reserve Requirements on Marginal Emission Rate. Arif, Md Arifin; Wang, Fengyu; Shi, Di; Sun, Liang; Wang, Zongjie (2024). 1–5.
• LDS2AE: Local Diffusion Shared-Specific Autoencoder for Multimodal Remote Sensing Image Classification with Arbitrary Missing Modalities. Qu, Jiahui; Yang, Yuanbo; Dong, Wenqian; Yang, Yufei M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 14731–14739.
• Learning Multi-Modal Cross-Scale Deformable Transformer Network for Unregistered Hyperspectral Image Super-resolution. Dong, Wenqian; Xu, Yang; Qu, Jiahui; Hou, Shaoxiong M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 1573–1581.
• Deep Reinforcement Learning-Enabled Adaptive Forecasting-Aided State Estimation in Distribution Systems with Multi-Source Multi-Rate Data. Zhang, Ying; Zhao, Junbo; Shi, Di; Chung, Sungjoo (2024). 1–5.
• Gradient-Guided Modality Decoupling for Missing-Modality Robustness. Wang, Hao; Luo, Shengda; Hu, Guosheng; Zhang, Jianguo M. J. Wooldridge, J. G. Dy, S. Natarajan (eds.) (2024). 15483–15491.
• TRIZ-GPT: An LLM-augmented method for problem-solving. Chen, Liuqing; Song, Yaxuan; Ding, Shixian; Sun, Lingyun; Childs, Peter R. N.; Zuo, Haoyu (2024).
• A Multi-Action Deep Reinforcement Learning Based on BiLSTM for Flexible Job Shop Scheduling Problem with Tight Time. Wang, Rui; Liu, Chang; Wang, Xinzhuo; Yang, Shengxiang; Hou, Yaqi X. Chen, J. Fang (eds.) (2024). 318–326.
• Insect diversity for agroecosystem resilience in a changing climate. Steffan-Dewenter, Ingolf; Bezner Kerr, Rachel; Peters, Marcell Karl (2024). 7(4) 541–544.

  Insect diversity plays a pivotal role for food security through crop damage, natural pest control, and pollination. Climate warming increases risks of insect pest outbreaks and threatens regulating services due to spatial and temporal mismatches. Transformation toward regionally adapted agroecological management of agroecosystems is urgently needed to support resilient food systems.
• Deep Active Learning Image Classification Algorithm Based on Class-Wise Self-Knowledge Distillation. Pang, Yuliang; Liu, Ying; Hao, Yu; Gong, Yanchao; Li, Daxiang; Xu, Zhijie (2023). 205–211.
• Edge-Computing Based Dynamic Anomaly Detection for Transmission Lines. Wang, Xinan; Shi, Di; Xu, Guangyue; Wang, Fengyu (2023). 1–5.
• A Game-Theoretic Access Strategy for Satellite Edge Computing enabled Massive IoT Networks. Liang, Hui; Zhang, Wei (2023). 5330–5335.
• An End-to-End Multi-stage Network for Ultrasound Video Object Segmentation. Wang, Mei; Liu, Shiyun; Dong, Yijie; Xu, Zhijie; Pan, Qiao; Chen, Dehua; Su, Jianwen X. Jiang, H. Wang, R. Alhajj, X. Hu, F. Engel, M. Mahmud, N. Pisanti, X. Cui, H. Song (eds.) (2023). 3574–3581.
• Feature Fusion-based Automatic Grading of Color Doppler Flow within Ovarian Masses in Ultrasound Images. Wang, Xu; Zhu, Linlin; Tao, Guowei; Ma, Xin (2023). 243–248.
• Cycle-Consistent Learning for Weakly Supervised Semantic Segmentation. Wang, Bin; Qiao, Yu; Lin, Dahua; Yang, Stephen D. H.; Li, Weijia D. Zhang, C. Fang, W. Liu, X. Liu, J. Song, H. Zhu, W. Huang, J. Smith (eds.) (2022). 7–13.
• SepFormer: Selective Patch Transformer for Image Classification. Zhang, Yang; Zhang, Zhaogong (2022). 69–74.
• FCOS Small Target Detection Algorithm Combined with Multi-Layer Hybrid Attention Mechanism. Liu, Ying; Geng, Luyao; Hao, Yu; Xu, Zhijie (2021). 50–55.
• Text Recognition in UAV Aerial Images. Wang, Shu; Wang, Dianwei; Han, Pengfei; Ren, Xincheng; Xu, Zhijie (2021). 232–238.
• An Unmanned Aerial Vehicle Video Object Tracking Algorithm Based on Siamese Attention Network. Zheng, Yuhuan; Wang, Dianwei; Han, Pengfei; Ren, Xincheng; Xu, Zhijie (2021). 1–8.
• Micromotion minimisation by synchronous detection of parametrically excited motion Nadlinger, D. P.; Drmota, P.; Main, D.; Nichol, B. C.; Araneda, G.; Srinivas, R.; Stephenson, L. J.; Ballance, C. J.; Lucas, D. M. (2021).
• An Anchor-free Detector Based on Residual Feature Enhancement Pyramid Network for UAV Vehicle Detection. Xie, Jianghuan; Wang, Dianwei; Guo, Jiaxing; Han, Pengfei; Fang, Jie; Xu, Zhijie (2021). 287–294.
• A No-reference Image Quality Assessment Method for Real Foggy Images. Wang, Dianwei; Zhai, Jing; Han, Pengfei; Jiang, Jing; Ren, Xincheng; Qin, Yongrui; Xu, Zhijie (2020). 120–125.
• Theoretical and Experimental Study of z-axis Acceleration Detection of the Micro Thermal Convective Accelerometer. Wang, Xiaoyi; Xu, Wei; Luo, Huahuang; Zhao, Xu; Lee, Yi-Kuen (2019). 1–4.
• Sensitivity Improvement of Micro Thermal Convective Accelerometer with Structure Optimization: Theoretical and Experimental Studies. Wang, Xiaoyi; Lim, Gyuha; Xu, Wei; Lee, Yi-Kuen (2019). 1–4.
• Adaptive Applications of Maximum Entropy Principle. Singh, Amit Kumar; Senapati, Dilip; Mukherjee, Tanmay; Rajput, Nikhil Kumar in Advances in Intelligent Systems and Computing, C. R. Panigrahi, B. Pati, P. Mohapatra, R. Buyya, K.-C. Li (eds.) (2019). (Vol. 1198) 373–379.
• A Non-stationary Analysis of Erlang Loss Model. Singh, Amit Kumar; Senapati, Dilip; Bebortta, Sujit; Rajput, Nikhil Kumar in Advances in Intelligent Systems and Computing, C. R. Panigrahi, B. Pati, P. Mohapatra, R. Buyya, K.-C. Li (eds.) (2019). (Vol. 1198) 286–294.
• A Neuron-Network-Based Optimal Control of Ultra-Capacitors with System Uncertainties. Duan, Jiajun; Yi, Zhehan; Shi, Di; Xu, Hao; Wang, Zhiwei (2019). 1–5.
• A Rprop-Neural-Network-Based PV Maximum Power Point Tracking Algorithm with Short-Circuit Current Limitation. Cui, Yao; Yi, Zhehan; Duan, Jiajun; Shi, Di; Wang, Zhiwei (2019). 1–5.
• Micro Humidity Sensor Using ALD Zirconium Oxide for Low Frequency Applications. Izhar; Wang, Xiaoyi; Xu, Wei; Lee, Yi-Kuen (2019). 1–4.
• Parametric Analysis and Efficient Optimization of Micro Thermal Convective Accelerometers with Excellent Sensitivity Using a Validated 1-D Model. Wang, Xiaoyi; Xu, Wei; Lijin, Beiqi; Ahmed, Moaaz; Bermak, Amine; Lee, Yi-Kuen (2018). 245–248.
• Graph Computation and Its Applications in Smart Grid. Liu, Guangyi; Liu, Kewen; Shi, Di; Zhu, Wendong; Wang, Zhiwei; Chen, Xi G. Karypis, J. Zhang (eds.) (2017). 507–510.
• Integrating Extractive and Abstractive Models for Long Text Summarization. Wang, Shuai; Zhao, Xiang; Li, Bo; Ge, Bin; Tang, Daquan G. Karypis, J. Zhang (eds.) (2017). 305–312.
• A graph theory-based online keywords model for image semantic extraction. Wang, Jing; Xu, Zhijie S. Ossowski (ed.) (2016). 67–72.
• Abstraction Level Taxonomy of Programming Language Frameworks. Kahanwal, Brijender (2013). 3(4) 1–12.
• Sensitive, 3D micromotion compensation in a surface-electrode ion trap. Technical Report (Master thesis), Eltony, Amira M. PhD thesis, Massachusetts Institute of Technology. (2013).
• Illuminant Condition Matching in Augmented Reality: A Multi-Vision, Interest Point Based Approach. Bingham, Mark; Taylor, David; Xu, Zhijie; Gledhill, Duke E. Banissi, M. Sarfraz, J. Zhang, A. Ursyn, C.-J. Wong, M. W. M. Bannatyne, J. J. Zhang, L. H. San, M. L. Huang (eds.) (2009). 57–61.
• Video Volume Segmentation for Event Detection. Wang, Jing; Xu, Zhijie; Xu, Qian E. Banissi, M. Sarfraz, J. Zhang, A. Ursyn, C.-J. Wong, M. W. M. Bannatyne, J. J. Zhang, L. H. San, M. L. Huang (eds.) (2009). 311–316.
• The Design, Development and Evaluation of Virtual Reality Learning Environment for Numeracy Concepts Using 3D Virtual Manipulatives. Daghestani, Lamya F.; Ward, Robert D.; Xu, Zhijie; Al-Nuaim, Hana M. Sarfraz, E. Banissi, C.-J. Wong (eds.) (2008). 93–100.