Biology intern supporting Arcadis ecological assessments, permitting, and field surveys. Collecting environmental data and contributing to technical reporting across Ontario.
About the role
- Research Intern at Mila focusing on advanced AI projects and culturally relevant content creation methodologies in collaboration with engineering teams.
Responsibilities
- **Main responsibilities**
- Design, formalize and explore different strategies (explicit, implicit and structured) to influence model behavior (prompt rewriting, contextual instructions, RAG).
- Implement rigorous comparative evaluation protocols and define precise metrics to measure cultural relevance, response diversity and the overall usefulness of outputs.
- Iteratively test and compare the performance of developed strategies across multiple state-of-the-art models (e.g., GPT, Claude, Gemini).
- Build a "golden dataset" of culturally ambiguous or representative test queries covering domains such as film, music, media and literature.
- Develop simple interfaces, dashboards or interactive notebooks to compare models in parallel and facilitate easy annotation of results.
- Leverage and analyze real interactions from the MCP (Model Context Protocol) system to understand agents' responses to steering signals and turn these logs into research insights.
- Identify effective approaches, study trade-offs between local relevance and global usefulness, and produce a final research report with the potential to contribute to academic publications.
- Contribute to cutting-edge research on foundational models, language models or vision models and their industrial applications.
- Implement solutions and experiment with pre-trained foundational models.
- Adapt the latest VLM and LLM architectures, training techniques and evaluation pipelines for domain-specific applications.
- Propose and investigate innovative research directions to improve the team’s internal prototypes.
- Analyze system performance and contribute to iterative improvements through experimentation and testing.
Requirements
- **Candidate profile**
- Enrolled in a PhD program or near completion of a PhD in Computer Science, Machine Learning, Natural Language Processing (NLP) or a related technical field.
- Strong skills in experimental methodology, test protocol design and qualitative/quantitative model evaluation.
- Practical, hands-on experience with LLMs, RAG systems or agentic architectures.
- Proficiency in the Python programming language and ability to design a simple experimentation interface or tool (dashboards, interactive notebooks).
- Excellent ability to collaborate and bridge research and engineering teams.
- Currently a graduate student or recent graduate at the master’s or doctoral level in Computer Science, Applied Mathematics, Machine Learning, Computer Vision or a related technical field.
- Good understanding of machine learning, especially deep learning.
- Knowledge of foundational models and deep learning model architectures.
- Experience with programming languages, notably Python, and frameworks such as PyTorch.
- Research skills with the ability to stay up to date with the latest trends in AI and machine learning.
- Solid understanding of machine learning and computer vision.
- Excellent applied research skills, including problem definition, solution exploration, and analysis and presentation of results.
- Strong problem-solving skills and a passion for innovation.
- Intermediate proficiency in French and English, due to interactions you will have in the role with some of our partners, stakeholders or members of our anglophone academic community.
- **Additional assets**
- Demonstrated interest in issues of local cultural representation, diversity and the societal impact of AI (particularly in the Québécois and Francophone context).
- Familiarity with the MCP (Model Context Protocol) server ecosystem or other steering-layer infrastructures.
- Prior experience integrating commercial large model APIs (OpenAI, Anthropic, Google, etc.).
- Previous experience in Natural Language Processing (NLP) or vision models.
- Experience training/fine-tuning and evaluating vision-and-language models.
- Knowledge of recent multimodal alignment techniques and experience with large multimodal datasets (image-text-audio).
- Familiarity with limitations and challenges in vision-language modeling.
- Experience in rapid prototyping using AI platforms for model development and refinement.
- Exposure to generative AI tools and pipelines, including APIs for annotation and data curation (e.g., GPT-4), is a plus.
Benefits
- **We want to hear from you**
- At Mila, diversity matters to us. We value a fair, open and respectful work environment that embraces differences. We encourage anyone who wishes to contribute to an evolving, stimulating ecosystem and to help build and sustain a healthy, inclusive culture to apply.
- Please note that only selected candidates will be contacted.
- https://mila.quebec/fr/protection-de-la-vie-privee
Job title
Job type
Experience level
Salary
Not specified
Degree requirement
Tech skills
Location requirements
Report this job
Found something wrong with the page? Please let us know by submitting a report below.
General Utility Worker cleaning and sanitizing kitchenware, equipment, and lodge facilities at Aramark’s remote oilsands site. Supporting safe, orderly food - service operations near Fort McMurray.
Part - time retail shift leader supervising Canna Cabana cannabis store operations. Leading sales, customer service, staff direction, compliance, and inventory management.
IT fulfilment coordinator managing telecom - focused IT requests, procurement, asset tracking, vendors, and project financials. Supporting reliable technology services across The Salvation Army in Canada.
Part - time Retail Shift Leader supporting store operations at Canna Cabana, Canada’s largest cannabis retail chain. Supervising staff, customers, sales targets, KPIs, and inventory.
Personal Care Specialist supporting Cencora’s Canadian patient services and pharma commercialization business. Processing prior authorizations, insurance verification, financial assistance, and specialty pharmacy referrals.
Technical specialist optimizing precast refractory solutions for RHI Magnesita’s North American industrial customers. Supporting field assessments, manufacturing, trials, product development, and sustainability initiatives.
ERP Release Coordinator coordinating Infor M3 testing, documentation, and deployments. Supporting Winpak’s sustainable packaging operations across manufacturing and supply chain processes.
General Utility Worker maintaining cleanliness, sanitation, food stations, and inventory at LandSea’s remote BC workforce camps. Supporting safe, well - serviced accommodations for onsite crews.
Field Safety Specialist advising Grainger customers on safety improvements and MRO solutions. Coaching sellers and driving workplace safety implementations across British Columbia.