Co-op Researcher – LLM Post-Training & Agentic Coding
Summary
Co-op researcher focused on LLM post-training, agentic coding, and building intelligent agents for reasoning and tool use, involving GPU-optimized training/inference infrastructure and large-scale evaluations.
- Build intelligent agents for reasoning and tool use
- Collaborate to prototype and translate research to systems
- Conduct LLM post training research
- Design experiment and evaluation plans
- Develop LLM training and inference infrastructure
- Develop agent training approaches
- Generate tasks and data
- Implement and analyze large scale evaluations
- Improve distributed execution and resource utilization
- Optimize GPU post training workflows
