Description
The STELLAR project, funded by the Wellcome Trust and OpenAI, is seeking a Research Scientist to develop and validate Conceptor-based steering methods for large language models (LLMs), applied to realistic AI-powered patient simulations for mental health clinician training. Conceptors are a mathematical framework for precisely manipulating internal LLM representations, enabling continuous control over psychiatric symptom presentations that simple prompting cannot.
The position is at the NYU Stern School of Business in the Department of Technology, Operations, and Statistics. You will report to Joao Sedoc and will collaborate with other stakeholders as well as with an interdisciplinary team spanning ML, psychiatry, clinical speech, and public health.
We are looking for someone who wants to do careful, impactful work at the intersection of representation learning and clinical AI.
Responsibilities
Develop Conceptor-based steering methods for open-weight LLMs (e.g., GPT-OSS, Llama, Gemma) to simulate psychiatric symptom presentations with precise, composable control
Design and run evaluation experiments measuring clinical fidelity, demographic fairness, and composability of digital patient twins
Collaborate with Penn psychiatry and LDC teams on integrating acoustic-linguistic features from clinical interview recordings
Contribute to open-source releases of Conceptor libraries, evaluation frameworks, and model checkpoints
Publish research at top NLP, Information Systems, ML, and interdisciplinary journals and venues (TACL, ISR, MISQ, JMLR, ACL, EMNLP, NeurIPS, IMCL, ICLR, AAAI, JAMA, npj Digital Medicine)
Mentor graduate students and contribute to a collaborative, cross-site research environment
In compliance with NYC's Pay Transparency Act, the annual base salary range for this position is $61,500-$102,500. New York University considers factors such as (but not limited to) the specific grant funding and the terms of the research grant when extending an offer.
Qualifications
Required
PhD in NLP, information systems, machine learning, computer science, or a closely related field
Hands-on experience with LLM internals: activations, representations, fine-tuning, and/or mechanistic interpretability
Strong Python and PyTorch skills; comfortable working with large model weights
Track record of publishing peer-reviewed research
Preferred
Familiarity with steering vectors, activation engineering, or representation editing
Experience with clinical NLP, mental health language data, or speech biomarker research
Experience working with open-weight LLMs at scale (LoRA fine-tuning, inference optimization)
Interest in fairness, bias, and demographic invariance in language model outputs
Application Instructions
Apply via Interfolio and include the following:
a cover letter
a research statement describing research interests and accomplishments including research they would like to conduct (or have conducted) in AI and data-driven research
a C.V.
two letters of recommendation
Applications will be reviewed on a rolling basis until the position is filled. We aim to respond to all applicants within four weeks of submission.
Research Scientist (Prof Sedoc)
New York University · New York, NY, USA ·
- Pay:
- $61,500-$102,500/yr
- Job type:
- Full Time
About New York University
From the time of its founding in 1831 until 1973, New York University housed its library collections in various campus locations. A major gift from Elmer Holmes Bobst and his wife, Mamdouha, enabled the University to bring its various collections together in a new library on Washington Square South in 1973.
The striking, 12-story Elmer Holmes Bobst Library, designed by Philip Johnson and Richard Foster, is the flagship of a 10-library system that provides students and faculty members with access to the world’s scholarship. The system holds 5.8 million volumes and more than 2 million e-books.
Located at the heart of NYU’s Washington Square campus, Bobst Library is central to its intellectual life. When school is in session, Bobst receives 10,000 visits per day and more than two million visits annually.
Bobst Library offers approximately 3,000 seats for student study. Its facilities include the Avery Fisher Center for Music and Media, one of the world’s largest academic media centers, with media-ready collaborative studies and carrels and the multi-channel Feldstein Immersion Room, which provides one of the most dramatic listening experiences in New York.
Also in Bobst Library, the Digital Studio offers a leading-edge resource for faculty and student projects and promotes access to digital resources for teaching, learning, research, and arts events. The Data Services Lab is equipped with high performance computers for qualitative, quantitative, and geographical research. The Studio and the Lab are staffed by expert consultants from the Libraries and NYU IT.
Bobst Library is home to the Libraries’ Special Collections Center, opened in September 2019 and designed for research into and exhibition of the Libraries’ extraordinarily rich special collections. The new Center co-locates the collections of three formerly separate repositories: Fales Library, Tamiment Library, and the University Archives. The renowned Tamiment Collections document the history of labor, socialism, anarchism, communism, and American radicalism. The Robert F. Wagner Labor Archives hold the Jewish Labor Committee Archives and the historical records of more than 130 New York City labor organizations. The Fales Collections include the Fales Collection of English and American Literature, one of the most comprehensive fiction collections in the country. The Fales Collections also focus on food studies, avant-garde New York music and art, and American history. For example, the unique Downtown Collection documents the Downtown New York literary and arts scene from the 1970s to the present, focusing on postmodern writing and dance, performance art, and the downtown music scene. Supporting NYU’s comprehensive programs in nutrition and food studies, the Marion Nestle Food Studies Collection documents American food history with a focus on New York City; the Dalia Carmel Collection includes regional and national cookbooks from all over the world. The Special Collections Center also houses the University Archives, which comprise the historical records of NYU.
NYU’s special collections are supported by the Barbara Goldsmith Preservation and Conservation Department, whose staff conserves and preserves archival material in all media, including papyrus, paper, plastic, magnetic tape, film, and more. NYU Digital Library Technology Services (DLTS) processes, enables access to, and preserves digital materials from both the NYU community and collaborating partner organizations.
Since 2004, NYU has renovated seven of Bobst Library’s 12 library floors.