DeCLaRe Lab · Nanyang Technological University, Singapore

Join DeCLaRe Lab

Working Here

You will develop models, evaluations, and open artifacts around a clearly stated question, and build toward a direction of your own within the lab's agenda. If a theme on the research page raises a question you would want to spend years on, that is the right reason to write.

How Work Gets Done Here

Almost nothing here is written alone. A paper carries 5.6 authors on average — a median of five — and typically one or two are DeCLaRe members, with the rest from elsewhere. Reading the lab's 167 papers since 2019 as a set of collaboration types gives the shape of it; pick a relationship to trace it through the map.

DeCLaRe collaboration network The publication archive at the centre, connected to internal collaboration, work initiated in the lab, external collaborators, research themes, and academic and industry partners. The same figures are listed beside the map. 49.1% 84.4% 89.8% 70.1% 98.0% 21.3% Publicationarchive InternalDeCLaRecollaboration Initiatedin DeCLaRe Externalcollaborators Multi-themepapers Academicpartners Industrypartners

Percentages summarize papers from 2019 onward; DeCLaRe members include current members and alumni. Internal collaboration excludes the PI.

Open Positions

PhD Students

For applicants drawn to one of the six themes — safety, trustworthiness, multimodality, AI for Science, efficiency, or embodied AI. Publications help but are not required; engineering ability, research judgment, and persistence matter as much.

Postdoctoral Researchers

For researchers with a strong publication record and a clear agenda that connects to the lab's themes. Proposing a new direction is welcome when it links to that agenda.

Research Interns and Visiting Researchers

For students and collaborators who can contribute to a focused study, benchmark, research tool, or open-source release.

Backgrounds we commonly draw on: machine learning, natural language processing, computer vision, audio and speech, robotics, efficient learning, and AI safety and alignment.

How to Apply

Read the research slides and browse recent publications first, then email soujanya [dot] poria [at] ntu [dot] edu [dot] sg with your CV, a transcript if applicable, representative work, and a short note naming the theme or paper your interests connect to.

Keep it short and specific — a concrete note about one paper says more than a general statement of interest.