Max Mühlhäuser
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Max Mühlhäuser has a total of 378 co-authors. Here are the top co-authors by number of shared publications:
Jürgen Steimle
Max Planck Institute for Informatics, Saarbrücken, Germany
18 shared publicationsJochen Huber
MIT Media Lab & Singapore University of Technology and Design, Cambridge, MA, USA
14 shared publicationsDaniel Schreiber
Technische Universität Darmstadt, Darmstadt, DE
12 shared publications- Jussi Kangasharju
University of Helsinki, Finland
12 shared publications - Florian Müller
TU Darmstadt
10 shared publications Immanuel Schweizer
Telecooperation, Computer Science, TU Darmstadt, Hochschlustr. 10 - building S202, D-64289 Darmstadt, Germany
10 shared publications
Latest Publications
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2875-2887
FM4MC: Improving Feature Models for Microservice Chains—Towards More Efficient Configuration and Validation
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2875-2887
Open AccessFM4MC: Improving Feature Models for Microservice Chains—Towards More Efficient Configuration and Validation
- Open Access
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 14-28
No Needles Attached? Inferring Energy Metabolism Zones and Lactate Accumulation from Touchscreen Input
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 14-28
Open AccessNo Needles Attached? Inferring Energy Metabolism Zones and Lactate Accumulation from Touchscreen Input
- Open Access
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 642-654
Comfort Is in the Air: Investigating the Perceived Comfort of Rotational and Translational Mid-Air Interactions
- 16 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 642-654
Open AccessComfort Is in the Air: Investigating the Perceived Comfort of Rotational and Translational Mid-Air Interactions
- Open Access
- 9 September 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Human-Computer Interaction
- Vol. 9 (5) , 1-29
From Pegs to Pixels: A Comparative Analysis of the Nine Hole Peg Test and a Digital Copy Drawing Test for Fine Motor Control Assessment MHCI012
User interaction with digital systems requires Fine Motor Control (FMC), especially if the interfaces are complex or require high fidelity and fine-grained interactions. Despite its importance, Fine Motor Control is often overlooked in interactive system design, partly because of its complex assessment. Measuring changes in fine motor abilities due to prolonged use or fatigue currently requires repeated manual testing. This paper analyzes the concept of using the digital mobile devices’ input behavior to assess the user’s Fine Motor Control. For this, we show that Fine Motor Control can be assessed for touch and stylus-based interaction with a digital mobile system. We conducted a user study, where participants performed a Nine Hole Peg Test and a predefined Copy Drawing Test before and after exercises that affect fine motor skills. Based on this data, we investigated how metrics such as pressure, velocity, and entropy for touch and stylus input can be used to predict Fine Motor Control.
- 9 September 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Human-Computer Interaction
- Vol. 9 (5) , 1-29
Open AccessFrom Pegs to Pixels: A Comparative Analysis of the Nine Hole Peg Test and a Digital Copy Drawing Test for Fine Motor Control Assessment MHCI012
User interaction with digital systems requires Fine Motor Control (FMC), especially if the interfaces are complex or require high fidelity and fine-grained interactions. Despite its importance, Fine Motor Control is often overlooked in interactive system design, partly because of its complex assessment. Measuring changes in fine motor abilities due to prolonged use or fatigue currently requires repeated manual testing. This paper analyzes the concept of using the digital mobile devices’ input behavior to assess the user’s Fine Motor Control. For this, we show that Fine Motor Control can be assessed for touch and stylus-based interaction with a digital mobile system. We conducted a user study, where participants performed a Nine Hole Peg Test and a predefined Copy Drawing Test before and after exercises that affect fine motor skills. Based on this data, we investigated how metrics such as pressure, velocity, and entropy for touch and stylus input can be used to predict Fine Motor Control.
- Open Access
SafePath: Conformal Prediction for Safe LLM-Based Autonomous Navigation
Large Language Models (LLMs) show growing promise in autonomous driving by reasoning over complex traffic scenarios to generate path plans. However, their tendencies toward overconfidence, and hallucinations raise critical safety concerns. We introduce SafePath, a modular framework that augments LLM-based path planning with formal safety guarantees using conformal prediction. SafePath operates in three stages. In the first stage, we use an LLM that generates a set of diverse candidate paths, exploring possible trajectories based on agent behaviors and environmental cues. In the second stage, SafePath filters out high-risk trajectories while guaranteeing that at least one safe option is included with a user-defined probability, through a multiple-choice question-answering formulation that integrates conformal prediction. In the final stage, our approach selects the path with the lowest expected collision risk when uncertainty is low or delegates control to a human when uncertainty is high. We theoretically prove that SafePath guarantees a safe trajectory with a user-defined probability, and we show how its human delegation rate can be tuned to balance autonomy and safety. Extensive experiments on nuScenes and Highway-env show that SafePath reduces planning uncertainty by 77\% and collision rates by up to 70\%, demonstrating effectiveness in making LLM-driven path planning more safer.
Open AccessSafePath: Conformal Prediction for Safe LLM-Based Autonomous Navigation
Large Language Models (LLMs) show growing promise in autonomous driving by reasoning over complex traffic scenarios to generate path plans. However, their tendencies toward overconfidence, and hallucinations raise critical safety concerns. We introduce SafePath, a modular framework that augments LLM-based path planning with formal safety guarantees using conformal prediction. SafePath operates in three stages. In the first stage, we use an LLM that generates a set of diverse candidate paths, exploring possible trajectories based on agent behaviors and environmental cues. In the second stage, SafePath filters out high-risk trajectories while guaranteeing that at least one safe option is included with a user-defined probability, through a multiple-choice question-answering formulation that integrates conformal prediction. In the final stage, our approach selects the path with the lowest expected collision risk when uncertainty is low or delegates control to a human when uncertainty is high. We theoretically prove that SafePath guarantees a safe trajectory with a user-defined probability, and we show how its human delegation rate can be tuned to balance autonomy and safety. Extensive experiments on nuScenes and Highway-env show that SafePath reduces planning uncertainty by 77\% and collision rates by up to 70\%, demonstrating effectiveness in making LLM-driven path planning more safer.
- Open Access
- 2 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 85526-85542
Automated Insights Into GitHub Collaboration Dynamics
Today, GitHub™ is the most widely used platform for open-source software development. Large projects may comprise hundreds of distributed collaborators and thousands of GitHub “issues” (structured discussions). It includes basic support for dealing with issues, via “Pull Requests” (PRs)—document changes that can manually be defined to “close” them (i.e., they address and thereby conclude the issue discussion). Unresolved issues can pile up. For example, at the time of writing, the Kubernetes repository has almost 2000 open issues; finding which ones a PR might close is a hard task by itself. We address this by automatically identifying issue-PR relationships using language models (LMs) and leveraging Information Retrieval (IR) techniques. To foster further research, we contribute a carefully curated novel dataset called CodeConvo, reflecting the most influential open-source repositories for code development as well as some technical document repositories. We use this dataset to benchmark several state-of-theart (SOTA) non-proprietary models that show exceptional performance on the MTEB benchmark, as well as to train and evaluate the performance of Smart Insights into GitHub Issue-PR Relations (SIGIR), our tailored model. The best SIGIR model/data combination yields an average Mean Reciprocal Rank (MRR) above 0.7, around 20% higher than the best baseline performance. Notably, ablation studies revealed that knowledge transfer occurs not only between different programming languages but also between code and technical documents, albeit to a lesser extent. a We believe that these results are encouraging and that they can stimulate the practical application of LMs for taming the complexity of very large projects, in GitHub and beyond. The models and dataset of this study can be found at Huggingface.
- 2 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 85526-85542
Open AccessAutomated Insights Into GitHub Collaboration Dynamics
Today, GitHub™ is the most widely used platform for open-source software development. Large projects may comprise hundreds of distributed collaborators and thousands of GitHub “issues” (structured discussions). It includes basic support for dealing with issues, via “Pull Requests” (PRs)—document changes that can manually be defined to “close” them (i.e., they address and thereby conclude the issue discussion). Unresolved issues can pile up. For example, at the time of writing, the Kubernetes repository has almost 2000 open issues; finding which ones a PR might close is a hard task by itself. We address this by automatically identifying issue-PR relationships using language models (LMs) and leveraging Information Retrieval (IR) techniques. To foster further research, we contribute a carefully curated novel dataset called CodeConvo, reflecting the most influential open-source repositories for code development as well as some technical document repositories. We use this dataset to benchmark several state-of-theart (SOTA) non-proprietary models that show exceptional performance on the MTEB benchmark, as well as to train and evaluate the performance of Smart Insights into GitHub Issue-PR Relations (SIGIR), our tailored model. The best SIGIR model/data combination yields an average Mean Reciprocal Rank (MRR) above 0.7, around 20% higher than the best baseline performance. Notably, ablation studies revealed that knowledge transfer occurs not only between different programming languages but also between code and technical documents, albeit to a lesser extent. a We believe that these results are encouraging and that they can stimulate the practical application of LMs for taming the complexity of very large projects, in GitHub and beyond. The models and dataset of this study can be found at Huggingface.
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-18
A Qualitative Investigation of User Transitions and Frictions in Cross-Reality Applications
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-18
Open AccessA Qualitative Investigation of User Transitions and Frictions in Cross-Reality Applications
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-6
AEXL: Enhancing Path Prediction with Active Explainable Learning via Large Language Models
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessAEXL: Enhancing Path Prediction with Active Explainable Learning via Large Language Models
- Open Access
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 20-28
CLEAR-Command: Coordinated Listening, Extraction, and Allocation for Emergency Response with Large Language Models
Achref Doula, Bela Bohlender, Max Mühlhäuser, Alejandro Sanchez Guinea. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations). 2025.
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 20-28
Open AccessCLEAR-Command: Coordinated Listening, Extraction, and Allocation for Emergency Response with Large Language Models
Achref Doula, Bela Bohlender, Max Mühlhäuser, Alejandro Sanchez Guinea. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations). 2025.
- Open Access
- 1 January 2025
- Published by Privacy Enhancing Technologies Symposium Advisory Board in Proceedings on Privacy Enhancing Technologies
- Vol. 2025 (1) , 654-670
“It’s Not My Data Anymore”: Exploring Non-Users’ Privacy Perceptions of Medical Data Donation Apps
This paper contributes an in-depth investigation (N=24) of privacy perceptions in the context of medical data donation apps. Medical data donation refers to the act of voluntarily sharing medical data with research institutions, which plays a crucial role in advancing healthcare research and personalized medicine. To design effective medical data donation apps, we need to understand how privacy expectations affect people's willingness to use such apps. We focus on non-users—those who have no experience with medical data donation apps—because gaining a deeper understanding of their perceptions is essential for fostering the adoption of these apps. Our findings highlight the importance of trust, transparency, and anonymity as driving factors. Participants expressed a willingness to share highly sensitive medical data with the apps if they were assured of complete anonymity, yet criticism regarding the risks of de-anonymization was also raised. Based on our results, we identify privacy awareness issues, especially concerning data sensitivity. Additionally, we explain the differences between participants' privacy expectations and preferences and what existing medical data donation apps offer. Finally, we provide guidance for the development of future user-centric medical data donation apps.
- 1 January 2025
- Published by Privacy Enhancing Technologies Symposium Advisory Board in Proceedings on Privacy Enhancing Technologies
- Vol. 2025 (1) , 654-670
Open Access“It’s Not My Data Anymore”: Exploring Non-Users’ Privacy Perceptions of Medical Data Donation Apps
This paper contributes an in-depth investigation (N=24) of privacy perceptions in the context of medical data donation apps. Medical data donation refers to the act of voluntarily sharing medical data with research institutions, which plays a crucial role in advancing healthcare research and personalized medicine. To design effective medical data donation apps, we need to understand how privacy expectations affect people's willingness to use such apps. We focus on non-users—those who have no experience with medical data donation apps—because gaining a deeper understanding of their perceptions is essential for fostering the adoption of these apps. Our findings highlight the importance of trust, transparency, and anonymity as driving factors. Participants expressed a willingness to share highly sensitive medical data with the apps if they were assured of complete anonymity, yet criticism regarding the risks of de-anonymization was also raised. Based on our results, we identify privacy awareness issues, especially concerning data sensitivity. Additionally, we explain the differences between participants' privacy expectations and preferences and what existing medical data donation apps offer. Finally, we provide guidance for the development of future user-centric medical data donation apps.