13930 - Research Associate
Grade UE07: £41,064 - £48,822 per annum
College of Science and Engineering / School of Informatics
Full-time: 35 hours per week
Fixed-term: for 24 months
The Opportunity:
We invite applications for a Postdoctoral Research Associate in machine learning based in the School of Informatics, University of Edinburgh. The postholder will also be formally affiliated with the EPSRC-funded Hub in Generative AI and work with Drs Siddharth N. and Michael Gutmann as part of the Hub. This is an outstanding opportunity to conduct methodological research at the frontier of machine learning and to collaborate across a vibrant national network of leading universities and industry partners.
The scope of the project will be defined together with the candidate and tailored to their strengths and interests but will broadly focus on one or both of the following topics:
Mutual information estimation and maximisation for continuous and discrete variables, with application to cross-modal data analysis or experimental design and active learning. This work stream will build on papers [1, 2].
Probabilistic latent variable modelling with hierarchically structured continuous and discrete variables for more efficient and effective generative modelling [3,4] and uncertainty quantification, with application to diffusion models [5].
The overarching goal is to advance methodology and to explore their use in real-world problems in collaboration with Hub partners.
Neural Mutual Information Estimation with Vector Copulas, NeurIPS 2025,
Implicit Deep Adaptive Design: Policy-Based Experimental Design without Likelihoods, NeurIPS 2021
Autoencoding Conditional Neural Processes for Representation Learning, ICML 2024
Banyan: Improved Representation Learning with Explicit Structure, ICML 2025
On Designing Diffusion Autoencoders for Efficient Generation and Representation Learning.
The position includes funding for international travel, e.g., for attending conferences, visiting research collaborators, and disseminating research findings. The researcher will have access to the compute infrastructure available to the School of Informatics and the AI Hub.
We welcome both local (UK-resident) and international applicants. We warmly welcome qualified candidates from all backgrounds to apply and particularly encourage applications from underrepresented groups in the field. We are strongly committed to offering everyone an inclusive and non-discriminating working environment.
Your skills and attributes for success
Essential:
PhD (or near completion) in Machine Learning, AI, Statistics, Applied Mathematics, or a related field.
Research experience in at least one of: probabilistic machine learning, diffusion/flow-based models.
Proficiency in modern ML toolchains (e.g., PyTorch, JAX) and reproducible research practices.
A track record of high-quality publications, e.g. at ICML, NeurIPS, ICLR, AISTATS, ACL, EMNLP, CVPR, JMLR, Machine Learning, and computational statistics journals.
Excellent communication skills and a collaborative mindset.
Desirable:
Research experience in mutual information estimation and/or experimental design.
Research experience in energy-based models and/or hierarchical generative models.
Strong cross-disciplinary experience and expertise.
Strong software engineering practices (testing, benchmarking, packaging, CI).
This post is full-time (35 hours per week); however, we are open to considering flexible working patterns. We are also open to considering requests for hybrid working (on a non-contractual basis) that combines a mix of remote and regular on-campus working.
Contact details for enquiries:
For Dr Siddharth N.: [email protected]
For Dr Michael Gutmann: [email protected]
College of Science and Engineering / School of Informatics
Full-time: 35 hours per week
Fixed-term: for 24 months
The Opportunity:
We invite applications for a Postdoctoral Research Associate in machine learning based in the School of Informatics, University of Edinburgh. The postholder will also be formally affiliated with the EPSRC-funded Hub in Generative AI and work with Drs Siddharth N. and Michael Gutmann as part of the Hub. This is an outstanding opportunity to conduct methodological research at the frontier of machine learning and to collaborate across a vibrant national network of leading universities and industry partners.
The scope of the project will be defined together with the candidate and tailored to their strengths and interests but will broadly focus on one or both of the following topics:
Mutual information estimation and maximisation for continuous and discrete variables, with application to cross-modal data analysis or experimental design and active learning. This work stream will build on papers [1, 2].
Probabilistic latent variable modelling with hierarchically structured continuous and discrete variables for more efficient and effective generative modelling [3,4] and uncertainty quantification, with application to diffusion models [5].
The overarching goal is to advance methodology and to explore their use in real-world problems in collaboration with Hub partners.
Neural Mutual Information Estimation with Vector Copulas, NeurIPS 2025,
Implicit Deep Adaptive Design: Policy-Based Experimental Design without Likelihoods, NeurIPS 2021
Autoencoding Conditional Neural Processes for Representation Learning, ICML 2024
Banyan: Improved Representation Learning with Explicit Structure, ICML 2025
On Designing Diffusion Autoencoders for Efficient Generation and Representation Learning.
The position includes funding for international travel, e.g., for attending conferences, visiting research collaborators, and disseminating research findings. The researcher will have access to the compute infrastructure available to the School of Informatics and the AI Hub.
We welcome both local (UK-resident) and international applicants. We warmly welcome qualified candidates from all backgrounds to apply and particularly encourage applications from underrepresented groups in the field. We are strongly committed to offering everyone an inclusive and non-discriminating working environment.
Your skills and attributes for success
Essential:
PhD (or near completion) in Machine Learning, AI, Statistics, Applied Mathematics, or a related field.
Research experience in at least one of: probabilistic machine learning, diffusion/flow-based models.
Proficiency in modern ML toolchains (e.g., PyTorch, JAX) and reproducible research practices.
A track record of high-quality publications, e.g. at ICML, NeurIPS, ICLR, AISTATS, ACL, EMNLP, CVPR, JMLR, Machine Learning, and computational statistics journals.
Excellent communication skills and a collaborative mindset.
Desirable:
Research experience in mutual information estimation and/or experimental design.
Research experience in energy-based models and/or hierarchical generative models.
Strong cross-disciplinary experience and expertise.
Strong software engineering practices (testing, benchmarking, packaging, CI).
This post is full-time (35 hours per week); however, we are open to considering flexible working patterns. We are also open to considering requests for hybrid working (on a non-contractual basis) that combines a mix of remote and regular on-campus working.
Contact details for enquiries:
For Dr Siddharth N.: [email protected]
For Dr Michael Gutmann: [email protected]
