Research Scientist (AI Model Evaluation and Adaptation), IHPC

Agency for Science, Technology and Research (A*STAR); Singapore; On-site; SGD 60,000 - 100,000; Full time

Job summary

A leading research institution in Singapore seeks a Research Scientist to evaluate and adapt AI models for online trust and safety applications. Candidates should possess a PhD in relevant fields, with strong skills in machine learning, data analysis, and programming. The role emphasizes interdisciplinary collaboration and continuous improvement in a dynamic research environment.

Qualifications

• PhD degree required in relevant fields.
• Demonstrable independent research capabilities and a strong publication record.
• Hands-on experience in AI frameworks and tools.

Responsibilities

• Evaluate AI models and datasets for performance and robustness.
• Collaborate with cross-functional teams and stakeholders.
• Generate reports and communicate findings effectively.

Skills

Machine Learning

Deep Learning

Natural Language Processing

Computer Vision

Data Visualization

Human-Computer Interaction

Data Analysis

Education

PhD in Computer Science / Information Technology / Computer Engineering / Information Systems / Computational Social Science / Societal Computing / Human Computer Interaction

Tools

Python

Java

C/C++

MATLAB

Job description

The Centre for Advanced Technologies in Online Safety (CATOS; https://www.catos.sg) was established in 2023 to host Singapore's Online Trust and Safety (OTS) Programme, a national research programme that leads the advancement of whole-of-nation technology capabilities to monitor and tackle online harms. The Systems Engineering Pillar of CATOS focuses on translational research and development, including evaluating, testing and integrating research output into needle-moving applications. One of the major outcomes will be a technological platform with an integrated suite of deep tech OTS engines, which analyses various internet sites and platforms for fast-trending harmful online content, such as non-factual claims, deepfakes, and hateful and toxic content. Different combinations of OTS engines can then be integrated and adapted to address the specific requirements of various stakeholders.

This role will be responsible for systematically evaluating AI models and datasets to understand the strengths and weaknesses of existing tools in this field and identify ways to adapt promising tools to their intended applications. The work will create valuable knowledge to inform end-users and technology researchers, helping to ensure that the technologies from CATOS and its partners are of desirable performance and robustness for the intended use cases and applications.

The candidate will assume the following roles and responsibilities:

As a Research Scientist, contribute to the evaluation of emerging technologies and data underneath the technologies, which often require hands-on tech skills and data analytic / data science capabilities.

Actively review relevant tech landscape from both academics and industry, construct datasets, perform evaluation, generate reports, and communicate the methodology, findings, and key takeaways.

Collaborate with senior team members, stakeholders and cross-functional teams to incorporate requirements that may come from the evolving nature of the landscape.

Contribute to key CATOS events that are aimed at engaging the researchers, industry, public agencies, NGOs and the general public.

JOB REQUIREMENT:

APhD degree in Computer Science / Information Technology / Computer Engineering / Information Systems / Computational Social Science / Societal Computing / Human Computer Interaction or equivalent.

A compelling scientific publications track record and demonstrable independent research capabilities.

Hands-on experience in using machine learning, deep learning frameworks and models, especially expertise in Natural Language Processing, Computer Vision, Data Visualization, and Human-Computer Interaction.

Proficient with programming languages. (e.g., Python, Java, C/C++, MATLAB)

Experience and expertise in (Multi-Modal) Large Language Models are especially valued.

Cross-disciplinary research experiences are especially valued, especially those related to online trust and safety.

Good communication and interpersonal skills for working in a multicultural work environment.

Candidates with no experience may apply.

Our research and engineering work is highly interdisciplinary, agile, and pragmatic in nature.

We expect everyone to take ownership of their role, responsibilities, and tasks.

We encourage everyone to speak upwhen they have feedback or ideas for improvement.

We provide honest feedback to one another, which can be tough at times, but it fosters rapid learning.

We careabout each other'sprofessional growth and well-being.

When applying, please share at least two representative research projects or publications that you have developed independently or contributed to significantly.

Please note that only shortlisted candidates will be notified.