Reference:

Senior Researcher/PostDoc (m/w/d/x): Multilingual MI

Senior Researcher/PostDoc (m/w/d/x): Multilingual MI

Full Job Title - DE
Wissenschaftliche*r Mitarbeiter*in/Postdoktorand*in (m/w/d/x): Mehrsprachige mechanistische Interpretierbarkeit
Full Job Title - EN
Senior Researcher/PostDoc (m/w/d/x): Multilingual Mechanistic Interpretability
Area
Science
Department
Multilinguality and Language Technology
Location
Saarbrücken
Contact
Dr. Simon Ostermann
Contact Phone no
+49 681 85775 5310
Address
Deutsches Forschungszentrum für Künstliche Intelligenz GmbH (DFKI); Stuhlsatzenhausweg 3;Saarland Informatics Campus D 3_2;66123 Saarbrücken;
Email
Simon.Ostermann@dfki.de
Employment Category
Full time
Type of Contract
Temporary
Department Description - DE
3976
Department Description - EN
3977
Term (months)
36
Valid to
15.09.2026
Ref No
11388-2026

We seek a postdoctoral researcher (starting date: January 1st, 2027) to conduct the first systematic investigation of reasoning circuit structure in compact multilingual language models. This is a bilateral DFKI-Inria collaboration project funded by BMFTR (SLIMS), bridging mechanistic interpretability and multilingual NLP, addressing a critical gap: existing circuit analysis focuses on large English-centric models, leaving the question of cross-lingual and cross-scale circuit transfer largely unexplored. 

You will work closely with the ALMAnaCH team at Inria Paris (Prof. Benoit Sagot, Dr. Djamé Seddah) on controlled validation using backdoored model suites as ground truth. The project includes planned bilateral research visits and joint workshops, enabling direct collaboration with leading researchers in multilingual modeling and interpretability methodology. 

You will apply the full mechanistic toolkit (activation patching, sparse autoencoders, natural language autoencoders, transcoders) to understand whether reasoning circuits survive compression and translate across typologically diverse languages. Your work will directly inform the design of more capable multilingual models and advance interpretability methods for linguistically diverse systems. We value rigorous methodology and open discussion of results, regardless of whether they confirm initial hypotheses or reveal surprising findings. 

Importantly, you will work as part of a fully integrated team composed of members of another BMFTR-funded project, SAgA, ensuring that your circuit-level discoveries directly feed into safe agentic system design, creating a tight feedback loop between mechanistic analysis and safety architecture. 

The position is embedded in the Multilinguality and Language Technology (MLT) group under the direction of Dr. Marius Mosbach, with broader departmental support from Prof. Kristian Kersting (SAINT deparment, Darmstadt) and Prof. Verena Wolf (NMM department, Saarbrücken); and in close colaboration with the ALMAnaCH team at Inria Paris (Prof. Benoit Sagot, Dr. Djamé Seddah); the primary supervisor of the position is Dr. Simon Ostermann; the position will be situated in his team "Efficient and Explainable NLP"

Your tasks

  • Apply circuit analysis methods to identify and characterize reasoning circuits in small multilingual models 

  • Investigate circuit transfer across model scales (1B-7B) and across 20-30 typologically diverse languages 
  • Validate next-generation interpretability tools (NLAs, transcoders) in the multilingual, small-model regime 
  • Design and conduct controlled validation experiments using backdoored model suites with known ground truth 
  • Publish results and release trained interpretability artifacts (SAEs, NLAs, transcoders) openly 

Your qualifications

Required Background 

  • PhD in computer science, NLP, linguistics, or machine learning 
  • Established expertise in mechanistic interpretability (circuit analysis, activation patching, SAEs, or equivalent) 
  • Strong background in multilingual or low-resource NLP 
  • Proficiency in Python and deep learning frameworks 
  • Track record of publications in top venues (ACL, EMNLP, NeurIPS, ICML, or equivalent) 

 

Desirable Qualifications 

  • Experience with sparse autoencoders, transcoders, or natural language autoencoders 
  • Prior work on cross-lingual transfer or multilingual representation analysis 
  • Familiarity with typologically diverse language families 
  • Interest in AI safety and alignment 
  • Experience with evaluation infrastructure and benchmarking 

Your benefits

  • We offer competitive, market-based compensation and many other benefits (Urban Sports Club, corporate benefits, a subsidy for your Jobticket, and much more)
  • 3-year appointment with possibility of extension 
  • Significant computational resources (GPU access) 
  • Embedded international collaboration with Inria Paris ALMAnaCH team, including planned bilateral research visits and joint workshop participation 
  • Direct mentorship from leading researchers in mechanistic interpretability and multilingual NLP across institutions 
  • Active publication culture with strong support for dissemination in top venues 
  • A rigorous, methodologically grounded research environment with genuine space for critical discussion and scientific independence 
  • Supportive, respectful lab culture that values both scientific excellence and human well-being 
  • Opportunity to co-supervise PhD and Master's students 
  • Access to open-weight multilingual models and established benchmarking infrastructure 

The German Research Center for Artificial Intelligence (DFKI) has operated as a non-profit, Public-Private-Partnership (PPP) since 1988. DFKI combines scientific excellence and commercially-oriented value creation with social awareness and is recognized as a major "Center of Excellence" by the international scientific community. In the field of artificial intelligence, DFKI has focused on the goal of human-centric AI for more than 35 years. Research is committed to essential, future-oriented areas of application and socially relevant topics.

DFKI encourages applications from people with disability; DFKI intends to increase the proportion of female employees in the field of science and encourages women to apply for this position.

Back to vacancies
Deadline for applications:
Please feel free to contact us if you have any questions regarding this position:
Share vacancy: