Pytorch, Python, Sql, regression comparison pipelines, Memory, search code interpreters, data pipelines, LLM Agents, error attribution, evaluation-guided Agent optimization, sandboxes, Verification, human annotation, verl, self-refine tool, post-training of LLMs, Agent reliability analysis, reward model design, LLaMA-Factory, RAG retrieval systems, context engineering, vLLM, training and inference frameworks, uncertainty estimation, benchmark construction, Harness design, MCP, FSDP