CS PhD & Postdoc Roundup — May 2026

NLP & Computational Linguistics

PhD in Formal Methods for NLP — Leiden University (LIACS)

• PI(s): Dr. Gijs Wijnholds (primary), Prof. Dr. Suzan Verberne (co-supervisor)
• Institution: Leiden Institute of Advanced Computer Science (LIACS), Leiden University, Netherlands
• What the lab does: LIACS is Leiden University's AI and computer science research institute; the hosting faculty has over 2,300 staff and approximately 5,000 students.1
• Research direction: Applying formal methods — logic and symbolic approaches — to improve the efficiency, transparency, and interpretability of LLMs. The project sits at the intersection of symbolic-neural integration, multilingual NLP, and vision-language models.
• Salary: €3,059–€3,881 gross/month (Dutch Payscale P), 38-hour week
• Duration: 1-year initial contract + 3-year extension, expected start Fall 2026
• Deadline: June 26, 2026 (23:59 Amsterdam time)
• Apply: Online via academicpositions.com vacancy 248628 or the Leiden University job portal; contact Dr. Wijnholds for academic questions

PhD in Mechanistic Interpretability for LLM Security — CopeNLU, University of Copenhagen

• PI(s): Prof. Isabelle Augenstein (lead), Dr. Pepa Atanasova (co-supervisor)
• Institution: CopeNLU group, Department of Computer Science, University of Copenhagen; affiliated with the Pioneer Centre for AI, Denmark
• What the lab does: CopeNLU is a competitive NLP group ranked 2nd in Europe and 19th globally on CSRankings (2023 data).2 The group is funded by an ERC Starting Grant on explainable and robust fact-checking, and had 6 papers accepted at ACL 2026, 8 at EMNLP 2025, 3 at ACL 2025, and 4 at NAACL 2025.3
• Research direction: Mechanistic interpretability methods to reduce the susceptibility of LLMs to false-information attacks, as part of the project "A Mechanistic Framework for Mitigating the Susceptibility of LLMs to Learning False Information," funded by the Independent Research Foundation Denmark.
• Lab culture: "We are a very social and collaborative group, with weekly group meetings, biweekly reading groups, one-on-one supervision meetings, regular social events, research retreats, and more."4 There is also an opportunity to collaborate with NVIDIA as an academic collaborator.
• Duration: 3 years, fully funded; start September 2026
• Deadline: May 31, 2026
• Apply: employment.ku.dk/phd/?show=160571

Postdoc in LLM Security & Explainable AI — CopeNLU, University of Copenhagen

• PI(s): Prof. Isabelle Augenstein, Dr. Pepa Atanasova
• Institution: CopeNLU group, University of Copenhagen (same lab as above)
• Research direction: Characterize the spectrum of false information that LLMs produce across their life cycle, develop mitigation methods, and harden LLMs against false-information attacks. The work intersects NLP security and explainable AI.2
• Duration: 3 years, fully funded; start September 2026
• Deadline: May 31, 2026
• Apply: employment.ku.dk/faculty/?show=160569

PhD in Multimodal & Social Grounding of Speech Language Models — Inria Grenoble

• PI(s): Thomas Hueber (primary), Stéphane Lathuilière, Laurent Girin; with collaborators from developmental psychology (Mathilde Fort) and Tampere University (Okko Räsänen)
• Institution: Joint position at GIPSA-lab (CNRS / Université Grenoble Alpes, ComLearn team) and Inria Centre at Université Grenoble Alpes (RobotLearn team), France. Inria is the French national research institute for digital science and technology, employing approximately 2,600 people.5
• Research direction: Grounded Speech Language Models (SpeechLMs) trained from raw speech, investigating how multimodal and social interactions drive human language acquisition. Covers speech/NLP, computer vision, deep learning, robotics, and cognitive science.
• Salary: ~€2,300 gross/month (standard French doctoral salary)
• Duration: 3 years, fully funded by the DevAI&Speech Chair of the MIAI Cluster; start October 1, 2026
• Benefits: Subsidized meals, partial public transit reimbursement, 7 weeks annual leave + 10 RTT days, telework options
• Deadline: June 12, 2026
• Apply: jobs.inria.fr/public/classic/en/offres/2026-10080 or contact [email protected]

PhD in LLM Agents for Multimodal Learning & Reasoning — FBK + Università di Padova ⚡ Deadline May 20

• PI(s): NLP Research Group at Fondazione Bruno Kessler (FBK) — specific PI not named in the posting
• Institution: Fondazione Bruno Kessler (FBK), Trento, Italy, in collaboration with Università di Padova; PhD program in "Brain, Mind and Computer Science." FBK is a major Italian public research foundation specializing in information technology and digital content.6
• Research direction: Large language models and agents, multimodal AI, learning and reasoning.
• Deadline: May 20, 2026 at 13:00 CEST — three days from issue date
• Apply: Via the link in the FBK LinkedIn post; additional details at the linked FBK page. PI name not disclosed in the LinkedIn posting — contact FBK directly for clarification before applying.

PhD Research Fellow in AI Safety — University of Bergen

• PI(s): Prof. Jan Arne Telle (academic contact); Head of Department Inge Jonassen
• Institution: Department of Informatics, University of Bergen, Norway
• Research direction: Emergent behaviors and representation learning in LLMs; causal and mechanistic approaches to interpretability; frameworks for steering model behavior in safety-critical contexts.7
• Deadline: Open (no fixed deadline stated in the posting)
• Apply: Upload application at Jobbnorge (Norwegian job portal); contact [email protected] or [email protected] (HR)

Computer Vision

PhD in Geometric 3D Reconstruction — Lund University

• PI(s): Viktor Larsson (primary), Anders Heyden (co-supervisor)
• Institution: Division for Computer Vision and Machine Learning, Centre for Mathematical Sciences, Faculty of Engineering (LTH), Lund University, Sweden
• Research direction: Learning-based methods for 3D reconstruction and navigation from images and video that produce geometrically consistent outputs; combining epipolar geometry and homographies with physically informed neural networks.8
• Lab size: Approximately 20 doctoral candidates plus several senior researchers as of April 2026.
• Funding: Swedish Strategic Research Environment ELLIIT ("Learning Geometric Representations" project); monthly salary
• Duration: 4 years full-time (extendable to 5 years with up to 20% teaching duties); start September 1, 2026
• Deadline: May 19, 2026 — verify current status before applying9
• Apply: lu.varbi.com/en/what:job/jobID:924422/

PhD in Realistic Computer Vision — University of East Anglia

• PI(s): Prof. Graham Finlayson
• Institution: School of Computing Sciences, Colour & Imaging Lab (C&IL), University of East Anglia, Norwich, UK. C&IL currently has 10 PhD students and postdocs. International collaboration with the Computer Vision Center, Universitat Autònoma de Barcelona.
• Research direction: Computer vision algorithms producing realistic outputs — image relighting, underwater footage processing, combining AI/ML with classical imaging methods.10
• Funding: Home tuition fees + annual stipend £20,408 (2026/27 rate) + £2,000/year research training. International applicants must cover the Home/International tuition fee differential.
• Duration: 3 years; start October 1, 2026
• Deadline: July 2, 2026
• Apply: Via jobs.ac.uk reference FINLAYSONG_U26CMP

PhD / Research Associate in ML for Computer Vision — TU Dresden

• PI(s): Prof. Dr. Bjoern Andres, Chair of Machine Learning for Computer Vision
• Institution: Institute of Artificial Intelligence, TU Dresden, Germany
• Research direction: Curiosity-driven basic research in one of three tracks: (i) machine learning and image analysis, (ii) machine learning and combinatorial optimization, or (iii) machine learning and automated theorem proving.11
• Funding: Fully funded, salary grade E13 TV-L (German public sector scale)
• Deadline: June 1, 2026
• Apply: phdscanner.com — TU Dresden listing

Postdoc in ML/CV for Automated Plant Phenotyping — ETH Zurich (SDSC)

• PI(s): Hosted at Swiss Data Science Center (SDSC); collaboration with Prof. Achim Walter (Crop Science Group) and Prof. Nina Buchmann (Grassland Sciences), ETH Zurich
• Institution: Swiss Data Science Center (SDSC), ETH Zurich, Zurich, Switzerland. SDSC is a joint initiative of ETH Zurich and EPFL focused on data science research at the intersection of computational methods and application domains.
• Research direction: ML and computer vision for crop mixture phenotyping — foundation models, domain adaptation, 3D reconstruction from multi-modal field imaging, active learning, and plant trait estimation from image time series (PhenoMix Project).12
• Deadline: Not explicitly stated on the listing (posted May 7, 2026)
• Apply: academicpositions.com — ETH Zurich SDSC listing

Postdoc in Physics-Aware Generative AI — Inria

• PI(s): Raoul de Charette
• Institution: Inria (French National Institute for Research in Digital Science and Technology)
• Research direction: Physics-aware generative AI; computer vision; physical and uncertainty-aware AI.13
• Start: Fall 2026
• Deadline: Not stated — contact PI directly via LinkedIn for current status
• Apply: Via LinkedIn post or PI contact

ML Theory, Reinforcement Learning & Optimization

PhD in Causal Inference & Machine Learning — TU Delft ⚡ Deadline May 20

• PI(s): Jesse Krijthe (Assistant Professor, Computer Science, TU Delft), Nan van Geloven (Biostatistics, Leiden University Medical Center)
• Institution: Pattern Recognition and Bioinformatics Group, Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS), TU Delft, Netherlands. Part of the multi-disciplinary "Safe Causal Inference" consortium.
• Research direction: Data-driven methods to detect and mitigate violations of causal assumptions (positivity violations), incorporating assumption violations into causal models, and improving the trustworthiness of causal claims.14
• Deadline: May 20, 2026 — three days from issue date; apply immediately
• Apply: careers.tudelft.nl — TU Delft Causal ML

Postdoc in Min-Max Optimization — Johns Hopkins University

• PI(s): Prof. Nicolas Loizou
• Institution: Department of Applied Mathematics and Statistics, Johns Hopkins University, Baltimore, USA. JHU's Applied Math & Statistics department has longstanding strengths in mathematical optimization and data science.
• Research direction: Foundational and algorithmic aspects of min-max optimization with applications to ML, AI, and large-scale reinforcement learning — including stochastic min-max optimization, accelerated methods, alternating-update strategies, proximal-gradient-type algorithms, and efficient sampling schemes.15
• Salary: $63,460–$100,000/year
• Duration: 2 years, with renewal possibility; teaching load typically one course per year
• Deadline: June 15, 2026
• Apply: facultyjobs.jhu.edu/positions/details?postingId=A-184958-3

Postdoc in Reinforcement Learning / Online Learning — Politecnico di Milano (RL³ Lab)

• PI(s): Dr. Alberto Maria Metelli
• Institution: DEIB (Department of Electronics, Information and Bioengineering), Politecnico di Milano, Italy. The RL³ Lab is part of ELLIS Unit Milan and is one of the most active RL research groups in Italy, with 4 faculty, 1 postdoc, and 15 PhD students as of May 2026.16
• Research direction: Reinforcement learning and/or online learning, funded by a FIS2 Starting Grant (the Italian equivalent of an ERC Starting Grant). Both theory-heavy tracks (RL theory, bandits, online learning) and experiment-oriented tracks (empirical RL, PyTorch) are welcome.
• Duration: 2 years, fully funded, no mandatory teaching (optional)
• Deadline: Rolling — candidates hear back within 2 months
• Apply: Email [email protected], subject line: [Postdoc Application – YOUR_SURNAME]

PhD in Reinforcement Learning — University of Zurich (ALPI Lab)

• PI(s): Prof. Giorgia Ramponi
• Institution: ALPI Lab, University of Zurich, Switzerland
• Research direction: Multi-agent learning, multi-agent LLM systems, imitation learning and learning from human feedback, sequential decision-making under constraints.17
• Structure: Position starts as a Research Assistant, then transitions to a fully funded PhD
• Deadline: Rolling
• Apply: Email [email protected], subject: AlpiApplication2026; include CV, transcript, and a short motivational letter

Research Scientist / Postdoc in High-Dimensional Statistics & DL Theory — RIKEN AIP

• Institution: High-Dimensional Structure Theory Team, RIKEN Center for Advanced Intelligence Project (AIP), Komaba Campus, University of Tokyo, Japan. RIKEN AIP is Japan's flagship AI research organization.18
• Research direction: High-dimensional statistics, deep learning theory, complex data analysis, optimal transport theory, and causal inference; constructing theoretical frameworks for high-dimensional data structures using mathematical statistics, probability theory, and statistical mechanics.
• Salary: Postdoc ¥382,610+/month; Research Scientist ¥448,990+/month (including discretionary work allowance)
• Contract: 1-year fixed-term, renewable up to 5 years (Postdoc) or 7 years (Research Scientist)
• Deadline: Rolling — apply as early as possible
• Apply: Email [email protected]

Postdoc in ML Theory, RL & Privacy-Preserving Learning — Microsoft Research

• Institution: Machine Learning Research Group, Microsoft Research. MSR runs an active basic-research program across theory, RL, privacy, and systems, distinct from product engineering roles.19
• Research direction: ML theory, reinforcement learning, privacy-preserving learning, and dynamical learning. Curiosity-driven projects are encouraged.
• Duration: 2-year contract, starting July 2026, ending June 30, 2028
• Contacts by area: ML Theory / Privacy / Dynamics → Huishuai Zhang ([email protected]); RL → Li Zhao ([email protected]); General → Jiang Bian ([email protected])
• Deadline: Rolling
• Apply: Contact the relevant area contact directly

Post-Doctoral Researcher — Google DeepMind India (Bengaluru)

• Institution: Google DeepMind, Bengaluru, Karnataka, India. Google DeepMind is Alphabet's AI research division, known for AlphaFold, Gemini, and reinforcement learning research.
• Research direction: Fundamental and applied research in computing, including human-centered AI, AI for social good, and research applicable to Google products. Candidates are expected to have a publication record at NeurIPS, CVPR, ICCV, or ICLR.20
• Duration: 12–24 months, fixed-term contract; 2026 start
• Deadline: Rolling
• Apply: Google Careers — DeepMind India postdoc

PhD in Probabilistic Machine Learning — Technical University of Denmark (DTU)

• Institution: Technical University of Denmark (DTU), Copenhagen, Denmark. DTU is a leading technical university in Scandinavia with strong programs in data science and ML.21
• Research direction: Probabilistic machine learning and decision-making under uncertainty; developing robust sequential decision-making methods using the active inference framework, with application scenarios including dynamic pricing.
• Duration: 3 years, fully funded; preferred start September 1, 2026 (negotiable)
• Requirements: 2-year MSc (120 ECTS), strong background in probabilistic ML and deep learning, proficiency in Python/PyTorch/JAX or TensorFlow
• Deadline: Not stated — rolling; contact the DTU group directly
• Apply: Via jobsinenglish.dk — DTU probabilistic ML listing

GenAI, Multimodal AI & AI Safety

PhD in Generative Modelling for Biomedical Discovery — University of Cambridge

• PI(s): Dr. Hana Aliee
• Institution: CRUK (Cancer Research UK) Cambridge Institute, University of Cambridge, UK. The CRUK Cambridge Institute is a world-class cancer research center embedded within the university.22
• Research direction: Probabilistic generative models (diffusion models, flow matching, self-supervised and autoregressive models), causal ML, graph neural networks, and dynamical systems (neural ODEs/SDEs) for multimodal biomedical data integration. The lab's stated aim is "to advance the theory of representation learning and causal reasoning in high-dimensional multimodal data."
• Funding: Full University fees + annual stipend starting at £22,500 (2026–27 rate), 4 years
• Eligibility: UK and overseas; First or Upper Second Class degree required
• Start date: January 2027
• Deadline: June 8, 2026
• Apply: postgraduate.study.cam.ac.uk — reference SW49526

Postdoc in LLM-Based Decision Support for Multimodal Situation Awareness — Chalmers University

• Institution: Division of Computing Science, Department of Computer Science and Engineering, Chalmers University of Technology, Gothenburg, Sweden. Chalmers is one of Sweden's top technical universities, ranked among the top 100 globally in engineering and technology.23
• Research direction: AI-powered recommendation engine for real-time emergency response: LLMs integrating live sensor data (drones, wearables, thermal imaging) with 3D digital twins of buildings; explainable AI for emergency commanders.
• Requirements: Doctoral degree, ML expertise, Python/PyTorch
• Deadline: June 1, 2026
• Apply: academicpositions.com — Chalmers LLM Decision Support

PhD in Context-Aware AI for Insect Monitoring — KU Leuven

• PI(s): Prof. Wouter Saeys, Dr. Bart De Ketelaere (MeBioS division)
• Institution: MeBioS (Mechatronics, Biostatistics and Sensors) division, KU Leuven, Belgium. KU Leuven consistently ranks among the most innovative universities in Europe.24
• Research direction: Context-aware and data-efficient AI for predictive insect monitoring in agriculture; computer vision combined with spatial and ecological context; representation learning and active learning for integrated pest management.
• Duration: 4 years, fully funded
• Deadline: June 30, 2026
• Apply: Via the form linked in the PhDScanner tweet

MATS Autumn 2026 Research Fellowship (AI Safety)

• Organization: MATS Program (Machine Learning Alignment Theory Scholars), a non-profit research training program
• Location: Berkeley, CA and London, UK
• Program dates: September 28 – December 4, 2026 (10 weeks), with 6–12 month extension option
• Mentors: Researchers at Anthropic, Google DeepMind, OpenAI, Redwood Research, and ARC25
• Research tracks: Empirical AI Safety (LLM safety, scalable oversight, model internals), Theory, Strategy & Forecasting, Policy & Governance, System Security, Biosecurity, Founding & Field-Building
• Note: This is a research fellowship, not a traditional PhD/postdoc position. Fellows often co-author papers and receive direct mentorship from lab researchers. It functions as an entry point to long-term AI safety research roles.
• Deadline: June 7, 2026
• Apply: forms.matsprogram.org/apply-a26

PhD in AI & Societal Impacts — Imperial College London (Frontier AI Research Lab)

• PI(s): Dr. Jonathan Richard Schwarz — previously a Research Fellow at Harvard and Research Scientist at Google DeepMind (PhD from UCL/Gatsby Unit); currently Head of AI Research at Thomson Reuters and Professor at Imperial College London
• Institution: Frontier AI Research Lab, Imperial College London, UK
• Research direction: AI governance, ethics, misinformation, political science, computational social science — work at the intersection of AI and societal impacts. Preferred background: social sciences combined with computer science.26
• Deadline: Not stated — contact PI directly
• Apply: Contact Dr. Schwarz directly via the announcement on X

PhD / Postdoc in Genomics & Machine Learning — NUS (Boxiang Liu Lab)

• PI(s): Dr. Boxiang Liu, Presidential Young Professor; PhD from Stanford University
• Institution: National University of Singapore (NUS), Singapore. NUS is consistently ranked among the top universities in Asia and globally.27
• Research direction: Single-cell genomics, splicing quantitative trait loci (sQTLs), and machine learning for biomedicine. The lab's ISSAC method achieves a 1.4–2.5× power increase over existing single-cell sQTL discovery approaches.
• Funding: Competitive salaries; wetlab experience is a plus
• Deadline: Rolling
• Apply: Email [email protected]

Postdoc in AI for Science / Scientific ML — Oak Ridge National Laboratory

• Institution: Oak Ridge National Laboratory (ORNL), Tennessee, USA. ORNL is the US Department of Energy's largest science and energy laboratory, with major capabilities in scientific computing and AI for science.28
• Research direction: Scientific machine learning — novel algorithms and architectures for scientific computing applications.
• Deadline: Not stated — rolling
• Apply: jobs.ornl.gov — ORNL AI for Science postdoc

PhD in Generative AI Security & Trustworthy ML — Virginia Commonwealth University

• PI(s): Dr. Syed Irfan Ali Meerza
• Institution: Virginia Commonwealth University (VCU), Richmond, Virginia, USA. VCU is a public research university with a growing AI research program.29
• Research direction: Generative AI security (copyright protection, misuse prevention), trustworthy and robust machine learning, AI for healthcare and sensing, adversarial machine learning environments.
• Start date: Fall 2026
• Deadline: Not stated — contact PI directly
• Apply: Email smeerza@vols.utk.edu with CV, transcripts, and a research statement

PhD in AI Systems & Human-AI Collaboration — Carnegie Mellon University (Next Cycle)

• PI(s): Dr. Shreya Shankar — incoming Assistant Professor, CMU Computer Science Department and Human-Computer Interaction Institute (by courtesy); previously PhD from Berkeley EECS, undergrad at Stanford CS; created DocETL and ZenBase
• Institution: Carnegie Mellon University, School of Computer Science, Pittsburgh, USA. CMU's CS department is consistently ranked top-3 globally.
• Research direction: AI systems, human-AI collaboration, databases and HCI.30
• Timeline: Dr. Shankar joins CMU in Fall 2027. Applications for this cohort are expected due around December 2026.
• Note: This opening is for the next application cycle — included here as an early signal for candidates planning Fall 2027 applications.
• Apply: List Dr. Shankar in your CMU application when the cycle opens; she has stated: "If you're interested in AI systems or human-AI collaboration, list me in your application."

Deadline snapshot