Home
Darmstadt

Location: Darmstadt

Bhavyajeet Singh is a PhD student and doctoral researcher at the Ubiquitous Knowledge Processing Lab, Technical University of Darmstadt, advised by Prof. Iryna Gurevych. His research focuses on developing NLP methods for cybersecurity applications like penetration testing and secure code generation. Previously, he was a Research Fellow at Microsoft Research India, where he worked on low-latency memory-augmented retrieval systems. This work has been published at ICML 2026 and was integrated into Bing Ads production pipelines. His broader research experience spans secure code generation, retrieval-augmented and memory-augmented NLP systems, and cross-lingual fact-to-text generation. One of his most influential earlier publications is XAlign: Cross-lingual Fact-to-Text Alignment and Generation for Low-Resource Languages, Tushar Abhishek, Shivprasad Sagare, Bhavyajeet Singh, Anubhav Sharma, Manish Gupta, and Vasudeva Varma, 2022. The work introduced a large-scale cross-lingual fact-to-text dataset with 0.45M pairs across eight low-resource languages, supporting research on grounded multilingual generation and cross-lingual alignment. The paper has gathered numerous citations, with the dataset being used my multiple downstream projects and models.

Bhavyajeet Singh is a PhD student and doctoral researcher at the Ubiquitous Knowledge Processing Lab, Technical University of Darmstadt, advised by Prof. Iryna Gurevych. His research focuses on developing NLP methods for cybersecurity applications like penetration testing and secure code generation.

Previously, he was a Research Fellow at Microsoft Research India, where he worked on low-latency memory-augmented retrieval systems. This work has been published at ICML 2026 and was integrated into Bing Ads production pipelines. His broader research experience spans secure code generation, retrieval-augmented and memory-augmented NLP systems, and cross-lingual fact-to-text generation.

One of his most influential earlier publications is XAlign: Cross-lingual Fact-to-Text Alignment and Generation for Low-Resource Languages, Tushar Abhishek, Shivprasad Sagare, Bhavyajeet Singh, Anubhav Sharma, Manish Gupta, and Vasudeva Varma, 2022. The work introduced a large-scale cross-lingual fact-to-text dataset with 0.45M pairs across eight low-resource languages, supporting research on grounded multilingual generation and cross-lingual alignment. The paper has gathered numerous citations, with the dataset being used my multiple downstream projects and models.

Google Scholar

Personal Website

MSc Eshetu Moges Abuhay is a Senior Technical Assistant at the University of Gondar and an incoming ELIZA Master’s scholar at Technische Universität Darmstadt. He previously graduated with top honors from Samara University (BSc) and the University of Gondar (MSc), where he led various technical initiatives. His research focus centers on explainable machine learning and data-driven intelligent systems. Recently, he developed advanced software and predictive models aimed at improving healthcare workflows. His most influential research is his thesis, “Data-Driven Prediction of Emergency Severity Index in Ethiopia Using Explainable Machine Learning” (2026), currently under review by the Emergency Medicine Journal. This work addresses critical mistriage issues by identifying key clinical predictors and quantifying minimum data requirements for accurate emergency decisions. Alongside his academic contributions, he provides professional service in hardware and software maintenance and was awarded the prestigious ELIZA Master’s scholarship for academic excellence.

MSc Eshetu Moges Abuhay is a Senior Technical Assistant at the University of Gondar and an incoming ELIZA Master’s scholar at Technische Universität Darmstadt. He previously graduated with top honors from Samara University (BSc) and the University of Gondar (MSc), where he led various technical initiatives.
His research focus centers on explainable machine learning and data-driven intelligent systems. Recently, he developed advanced software and predictive models aimed at improving healthcare workflows. His most influential research is his thesis, “Data-Driven Prediction of Emergency Severity Index in Ethiopia Using Explainable Machine Learning” (2026), currently under review by the Emergency Medicine Journal. This work addresses critical mistriage issues by identifying key clinical predictors and quantifying minimum data requirements for accurate emergency decisions. Alongside his academic contributions, he provides professional service in hardware and software maintenance and was awarded the prestigious ELIZA Master’s scholarship for academic excellence.

Facebook

Hossein is a master’s student in Artificial Intelligence and Machine Learning at TU Darmstadt. He is currently a student research assistant at the Multimodal AI Lab, where he actively leads and contributes to top-tier research projects in his field. His current work focuses on multimodal generative models and their safety, with a particular interest in the personalization of generative vision models to make visual content creation more controllable, creative, and expressive. Looking ahead, Hossein intends to pursue a PhD centered on giving users intuitive control over image and video-synthesis models, as well as adapting large vision-language models for personalized tasks. Ultimately, he aspires to work at the intersection of AI, media, and entertainment.

Hossein is a master’s student in Artificial Intelligence and Machine Learning at TU Darmstadt. He is currently a student research assistant at the Multimodal AI Lab, where he actively leads and contributes to top-tier research projects in his field. His current work focuses on multimodal generative models and their safety, with a particular interest in the personalization of generative vision models to make visual content creation more controllable, creative, and expressive. Looking ahead, Hossein intends to pursue a PhD centered on giving users intuitive control over image and video-synthesis models, as well as adapting large vision-language models for personalized tasks. Ultimately, he aspires to work at the intersection of AI, media, and entertainment.

Google Scholar

I work where computer science meets the social sciences: I hold degrees in Computer Science, Peace and Conflict Research, and Philosophy, Politics and Economics. This combination reflects a lasting fascination with how technology and society shape one another and the question: how do we build AI systems that are both technically robust and trustworthy? I am interested in approaching this through disinformation detection and the study of how independent systems learn from and interact with one another, and through symbolic AI, reasoning, and causality, where my love of mathematics and formal logic meets my commitment to more transparent, reliable models. Moreover, I have research experience in cryptography, particularly updatable encryption and authentication, and I am keen to explore the intersections between cryptography and AI.

I work where computer science meets the social sciences: I hold degrees in Computer Science, Peace and Conflict Research, and Philosophy, Politics and Economics. This combination reflects a lasting fascination with how technology and society shape one another and the question: how do we build AI systems that are both technically robust and trustworthy? I am interested in approaching this through disinformation detection and the study of how independent systems learn from and interact with one another, and through symbolic AI, reasoning, and causality, where my love of mathematics and formal logic meets my commitment to more transparent, reliable models. Moreover, I have research experience in cryptography, particularly updatable encryption and authentication, and I am keen to explore the intersections between cryptography and AI.

I am a first-generation Computer Science graduate from India, currently pursuing a Master’s in Artificial Intelligence and Machine Learning at TU Darmstadt, driven by a deep commitment to building intelligent systems that are explainable, equitable, and impactful. My background uniquely bridges immersive 3D web engineering with emerging AI research, giving me both the engineering depth to implement real-world machine learning systems and the creative perspective to design human-centered AI experiences.

I am a first-generation Computer Science graduate from India, currently pursuing a Master’s in Artificial
Intelligence and Machine Learning at TU Darmstadt, driven by a deep commitment to building intelligent
systems that are explainable, equitable, and impactful. My background uniquely bridges immersive 3D web
engineering with emerging AI research, giving me both the engineering depth to implement real-world machine
learning systems and the creative perspective to design human-centered AI experiences.

I am currently pursuing my Master’s degree in Artificial Intelligence and Machine Learning at TU Darmstadt. Additionally, I am working on finishing a Bachelor’s degree in Mathematics. In the past, I have worked as a working student in Software Engineering at Bosch Automotive Steering. My past research experience has mainly focused on developing new reinforcement learning algorithms. For my Bachelor’s thesis, I worked on Gradient Iterated Temporal-Difference Learning under the supervision of Théo Vincent at the Intelligent Autonomous Systems group and DFKI and we are currently working on extending this work to Distributional RL. Gradient Iterated Temporal-Difference Learning was later published at the Reinforcement Learning Conference 2026 (https://arxiv.org/abs/2603.07833).

I am currently pursuing my Master’s degree in Artificial Intelligence and Machine Learning at TU Darmstadt. Additionally, I am working on finishing a Bachelor’s degree in Mathematics. In the past, I have worked as a working student in Software Engineering at Bosch Automotive Steering. My past research experience has mainly focused on developing new reinforcement learning algorithms. For my Bachelor’s thesis, I worked on Gradient Iterated Temporal-Difference Learning under the supervision of Théo Vincent at the Intelligent Autonomous Systems group and DFKI and we are currently working on extending this work to Distributional RL. Gradient Iterated Temporal-Difference Learning was later published at the Reinforcement Learning Conference 2026 (https://arxiv.org/abs/2603.07833).

Google Scholar

I am a PhD student in Multimodal Artificial Intelligence at TU Darmstadt under the supervision of Marcus and Anna Rohrbach and an ELLIS PhD candidate. My research focuses on trustworthy AI, with particular interests in adversarial robustness, AI security, multimodal learning, and vulnerabilities of unified and agent-based systems. I study how modern AI models can be manipulated and how such weaknesses can be detected and mitigated before deployment. Alongside my research, I lead teaching activities for the Statistical Machine Learning course at TU Darmstadt, one of the university’s largest machine learning courses with more than 500 students, and supervise student research projects. Before my PhD, I worked as an independent ML engineer and conducted more than 100 AI workshops across academia and industry. My most influential recent work is GEM: Geometric Erasure by Contrastive Velocity Matching in Rectified Flows (Grebe, Braun, Rohrbach & Rohrbach, ICML 2026 Spotlight), which introduces a novel geometric framework for robust concept erasure in generative models. I was also recognized as an ICML 2026 Gold Reviewer for outstanding reviewing contributions.

I am a PhD student in Multimodal Artificial Intelligence at TU Darmstadt under the supervision of Marcus and Anna Rohrbach and an ELLIS PhD candidate. My research focuses on trustworthy AI, with particular interests in adversarial robustness, AI security, multimodal learning, and vulnerabilities of unified and agent-based systems. I study how modern AI models can be manipulated and how such weaknesses can be detected and mitigated before deployment. Alongside my research, I lead teaching activities for the Statistical Machine Learning course at TU Darmstadt, one of the university’s largest machine learning courses with more than 500 students, and supervise student research projects. Before my PhD, I worked as an independent ML engineer and conducted more than 100 AI workshops across academia and industry. My most influential recent work is GEM: Geometric Erasure by Contrastive Velocity Matching in Rectified Flows (Grebe, Braun, Rohrbach & Rohrbach, ICML 2026 Spotlight), which introduces a novel geometric framework for robust concept erasure in generative models. I was also recognized as an ICML 2026 Gold Reviewer for outstanding reviewing contributions.

Google Scholar

Baraa Hikal is an M.Sc. student in Artificial Intelligence and Machine Learning at TU Darmstadt, an ELIZA scholarship holder, and a recipient of the Niedersachsenstipendium. He is currently a research assistant at the UKP Lab, working with Prof. Iryna Gurevych and Jonathan Tonglet on multimodal chronolocation, automated fact-checking, and planning agents that combine visual reasoning with external evidence retrieval. Previously, he worked at GippLab on dimensional sentiment analysis and deepfake analysis, and at Qatar University on IRAC-based legal summarization. His research focuses on trustworthy and agentic AI, multimodal reasoning, hallucination detection, and factuality evaluation. His bachelor’s thesis focused on hallucination detection in language models. He has achieved first-place results in shared tasks affiliated with ACL and EMNLP, including SemEval 2025, SemEval 2026, BEA 2025, and ArabicNLP 2025, and received the Best Paper Award at BEA 2025.

Baraa Hikal is an M.Sc. student in Artificial Intelligence and Machine Learning at TU Darmstadt, an ELIZA scholarship holder, and a recipient of the Niedersachsenstipendium. He is currently a research assistant at the UKP Lab, working with Prof. Iryna Gurevych and Jonathan Tonglet on multimodal chronolocation, automated fact-checking, and planning agents that combine visual reasoning with external evidence retrieval. Previously, he worked at GippLab on dimensional sentiment analysis and deepfake analysis, and at Qatar University on IRAC-based legal summarization. His research focuses on trustworthy and agentic AI, multimodal reasoning, hallucination detection, and factuality evaluation. His bachelor’s thesis focused on hallucination detection in language models. He has achieved first-place results in shared tasks affiliated with ACL and EMNLP, including SemEval 2025, SemEval 2026, BEA 2025, and ArabicNLP 2025, and received the Best Paper Award at BEA 2025.

Google Scholar

Principal Scientist at Merck Healthcare, part of Merck KGaA, Darmstadt.
Dr. sc ETH in Pharmaceutical Sciences, ETH Zurich.
Bachelor & Master in Chemistry at ETH Zurich, Switzerland.

Interested in AI for Drug Discovery and Development.

Google scholar

Justus Thies is a full professor for 3D Graphics & Vision at the Technical University of Darmstadt. Before joining TU Darmstadt, he was an independent research group leader at the Max Planck Institute for Intelligent Systems in Tübingen (2021-2024). He received his doctoral degree from the Friedrich Alexander University of Erlangen-Nuremberg in 2017 and worked as a postdoctoral researcher at TU Munich (2017-2021). He has a strong focus on marker-less motion capturing of facial performances, human bodies, and general non-rigid objects, as well as AI-based synthesis techniques that allow for photorealistic image and video synthesis. He was recently awarded with the German Pattern Recognition Award 2024, the Eurographics Young Researcher Award 2024 and an ERC Starting Grant 2024.

Justus Thies is a full professor for 3D Graphics & Vision at the Technical University of Darmstadt. Before joining TU Darmstadt, he was an independent research group leader at the Max Planck Institute for Intelligent Systems in Tübingen (2021-2024). He received his doctoral degree from the Friedrich Alexander University of Erlangen-Nuremberg in 2017 and worked as a postdoctoral researcher at TU Munich (2017-2021). He has a strong focus on marker-less motion capturing of facial performances, human bodies, and general non-rigid objects, as well as AI-based synthesis techniques that allow for photorealistic image and video synthesis. He was recently awarded with the German Pattern Recognition Award 2024, the Eurographics Young Researcher Award 2024 and an ERC Starting Grant 2024.

Personal website

Google scholar