Tasks:
- * Analyze model behavior and failure modes
- * Build agent evaluation systems
- * Design reward signals and graders
- * Develop credit assignment mechanisms
- * Develop data synthesis and filtering pipelines
- * Evolve models and agent harnesses collaboratively
- * Explore self-improvement and continual learning
- * Improve agent generalization and reliability
- * Improve reinforcement learning stability and efficiency
- * Research agentic reinforcement learning algorithms
Perks/Benefits:
Skills/Tech stack required:
[Agentic reinforcement learning] [AI Agents] [Code generation] [Credit Assignment] [Data Filtering] [Data Synthesis] [Evaluation Design] [Information Retrieval] [Large-scale] [Large-scale model] [Large-scale Model Training] [LLM post training] [Long Horizon Planning] [Model Evaluation] [Model Training] [Post-training] [Python] [PyTorch] [Reinforcement Learning] [Reward Modeling] [Tool use]
Educational requirements:
[Doctor of Philosophy] [Master's Degree]
Role(s):
[Agent Research Scientist] [AI Research Scientist] [Machine Learning Research Scientist] [Research Scientist] [Scientist]