Linköping University
Just Landed
Posted Yesterday
Postdoc in Causal Inference and Natural Language Processing at Linköping University Linköping University in Sweden
Degree Level
Postdoc
Field of study
Computer Science
Funding
Postdoctoral employment at Linköping University for two years with the possibility of extension up to a maximum of three years. Full-time position; salary is set individually according to collective agreements and university guidelines. Teaching may account for up to 20% of working hours.
Deadline
Sep 24, 2026
Country
Sweden
University
Linköping University
How do I apply for this?
Sign in for free to reveal details, requirements, and source links.
Apply for this position
Keywords
Suggested positions
About this position
Postdoc in Causal Inference and Natural Language Processing at Linköping University, Sweden.
This postdoctoral position sits at the intersection of causal inference, natural language processing, computational social science, and machine learning. The project, Countering Bias in AI Methods in the Social Sciences, is a collaboration between the Institute for Analytical Sociology (IAS) at Linköping University and the Department of Computer Science and Engineering (CSE) at Chalmers University of Technology, funded by WASP-HS. The role is based at IAS and involves close collaboration with the CSE team and partners at the University of Texas at Austin.
The research focuses on using text and other unstructured data for causal analysis while addressing treatment leakage and confounding in observational studies. Topics mentioned include large language models for removing treatment-predictive information from text, debiasing estimators such as Design-based Supervised Learning and Prediction-Powered Inference, and mechanistic interpretability methods such as sparse feature circuits and the SHIFT method. The applied side of the project includes estimating the impact of development projects in Africa, including IMF, World Bank, UNDP, and Chinese projects, using policy text and other modalities.
Applicants should hold a doctoral degree (or equivalent) in social sciences, computational social science, or a closely related quantitative field. Strong programming skills in Python or R are expected, together with excellent English. Training in either causal inference or NLP is essential, and experience with observational data, text as data, large language models, mechanistic interpretability, or semiparametric/design-based estimation is highly valued. The position is full-time and may include up to 20% teaching.
The appointment is a temporary contract for two years, with the possibility of extension up to a total of three years. Salary is individually set under applicable collective agreements. Applications must be submitted through the university’s online portal by 24 September 2026. Selected publications should be attached electronically; non-electronic publications must be mailed in three copies to the University Registrar so they arrive by the deadline.
Funding details
Postdoctoral employment at Linköping University for two years with the possibility of extension up to a maximum of three years. Full-time position; salary is set individually according to collective agreements and university guidelines. Teaching may account for up to 20% of working hours.
What's required
A doctoral degree or equivalent is required by the decision date, preferably awarded no more than three years before the application deadline unless special reasons apply. The ideal candidate has a PhD in social sciences, computational social science, or a closely related quantitative field such as statistics, computer science, machine learning, or natural language processing. Strong programming skills in Python or R are expected, along with excellent written and spoken English. Solid training in at least one of the two pillars, causal inference or natural language processing, and a genuine interest in bridging them are required. Experience with causal inference on observational data, text as data, large language models, mechanistic interpretability, or semiparametric/design-based estimation is a strong merit, as is a publication record in leading venues.
How to apply
Apply via the university’s online application portal by clicking the Apply button. Submit your application and selected research publications electronically before the deadline. If publications cannot be sent electronically, mail three sets to the University Registrar so they arrive by the deadline.
More information can be found here
Ask ApplyKite AI
How do I apply for this?
Sign in for free to reveal details, requirements, and source links.