Research Scientist (LLM Foundation Models)

Location: Hong Kong or Singapore. Remote-first, work-from-home arrangement; availability may vary depending on the role and team.

Company

Binance operates a global blockchain ecosystem covering cryptocurrency trading, finance, payments, education, research, and digital-asset services.

What you will do

• Advance reasoning and planning capabilities across foundation-model development, including data acquisition, evaluation, supervised fine-tuning, reward modelling, and reinforcement learning.
• Synthesize large-scale datasets through rewriting, augmentation, and generation for pretraining, SFT, and RL stages.
• Solve complex tasks using System 2 thinking and advanced decoding strategies such as MCTS and A*.
• Design and implement evaluation methodologies for foundation models at different development stages.
• Enable models to use external tools, APIs, and code interpreters, and build agents and multi-agent systems for complex tasks.

Requirements

• Research experience with reinforcement learning and large language models, with familiarity with large-scale model training.
• Strong data structures and fundamental algorithm skills.
• Fluency in Python, C++, or Java.
• Experience with influential projects or papers in reinforcement learning, NLP, or deep learning.
• Excellent problem analysis, problem-solving, communication, and collaboration skills.

Culture & Benefits

• Opportunity to work on challenging projects shaping the future of finance.
• Collaboration with global, user-focused teams in a flat organizational structure.
• Autonomy in a fast-paced and innovative environment.
• Career growth and continuous learning opportunities.
• Competitive salary and company benefits.