Research Internship : Reliable and Adaptive Machine Learning

Sigma Nova; Paris; On-site; EUR 8,900 - 13,000; Full time

Job summary

Sigma Nova invites a research intern to work in Paris from March to April 2027, exploring fundamental ML questions about adapting to new datasets, reliable uncertainty, and privacy-aware learning. You’ll join researchers to connect foundation models with probability, Bayesian inference, and optimal transport, shaping a topic with potential for publication.

The role blends mathematical thinking with hands-on experiments, offering mentorship, access to GPU resources, and room to influence the

Qualifications

• Master’s-level training in applied mathematics, statistics, ML, or related field.
• Strong Python skills and PyTorch experience.
• Solid foundations in probability, linear algebra, and optimization.
• Ability to read research papers and design reproducible experiments.
• Interest in Bayesian inference, optimal transport, or privacy-preserving learning.

Responsibilities

• Investigate how models adapt to new datasets.
• Explore how models provide uncertainty estimates with predictions.
• Study learning from multiple data sources while preserving privacy.
• Contribute to experiments and potential publications.
• Collaborate with researchers on probabilistic and transport-based methods.

Skills

Python

PyTorch

Probability

Linear algebra

Optimization

Reading papers

Reproducible experiments

Privacy-preserving learning

Education

Master's degree in applied mathematics, statistics, ML, or related field

Job description

Sigma Nova is looking for a research intern (March-April 2027) to investigate fundamental questions in machine learning: how can models adapt to new datasets, express reliable uncertainty, and learn from heterogeneous sources while respecting privacy?

This internship combines mathematical thinking with hands-on experimentation. You will work with researchers on methods that connect modern foundation models with ideas from probability, Bayesian inference, and optimal transport. The exact topic will be shaped around your interests and background, with a focused scope and opportunities to contribute to a publication.

Possible research tracks

1) Bayesian learning and uncertainty

Explore how models can provide useful uncertainty estimates alongside their predictions. Possible directions include scalable Bayesian approaches, parameter-efficient adaptation, and evaluating reliability when models encounter unfamiliar data.

2) Learning from distributed and heterogeneous data

Investigate how models can learn from multiple data sources without bringing all the data together. Topics may include federated learning, combining information across different populations, and understanding trade-offs between predictive performance, communication costs, and privacy.

3) Optimal transport and adaptation

Study how the geometry of probability distributions can help models compare datasets and transfer knowledge between tasks. Possible directions include learning compact dataset representations, selecting informative examples, and improving adaptation from limited observations.

These are alternative starting points rather than a single combined project. Depending on the topic, the balance may lean toward method development, mathematical analysis, or empirical evaluation.

Master’s-level training in applied mathematics, statistics, machine learning, or a related fieldStrong Python skills and experience with PyTorchSolid foundations in probability, linear algebra, and optimizationComfortable reading research papers and designing careful, reproducible experimentsInterest in Bayesian inference, optimal transport, or privacy‑preserving learningPrior experience in one of these areas is a plus; expertise in all of them is not expectedWhat we offerA research-driven startup working on foundation modelsClose mentorship from researchers working on optimal transport and Bayesian deep learningHigh-end GPU resources and existing research code to build onRoom to shape the research direction and contribute to a publication