Job Description  

Candidate will work on: 

Utilize open-source agents/models on new repositories to create more training data.  Evaluate various agents/models on existing well known SWE benchmarks.  Conduct post-training on collected training data to create fine-tuned models. Publish findings internally or externally (e.g., open-source repositories / benchmarks, academic conferences).  

Qualifications  

Currently pursuing a PhD in Computer Science, Software Engineering, Artificial Intelligence, Machine Learning, or a related field.   Strong research background in one or more of the following: LLM training in particular Supervised Fine Tuning or Reinforcement Learning, AI agents and autonomous systems, software engineering automation, etc.  Proficient in Python or Java.  Excellent analytical, problem-solving, and communication skills.   Experience with long-context model post-training (e.g., fine-tuning or evaluation for extended context windows) is a plus.  Publication record in top AI/ML/SE conferences