Uploaded October 2025 | Updated November 2026, 3 days ago
n all LLM-as-a-judge evaluations, two choices shape how reliable and useful the results are: whether you ask for explanations and whether you add chain of thought (CoT) prompting.
In this session, Elizabeth Hutton (Senior AI Engineer - Evals) and Sri Chavali (AI Engineer) will look at whether explanation order matters, when CoT is genuinely helpful, and why clear, well-structured prompts often outperform extra reasoning steps.
Learn practical strategies for making LLM evaluations more accurate, transparent, and scalable.
Notebook: github.com/Arize-ai/phoenix/blob/main/tutorials/evals/CoT_explanations_simple_vs_complex_evals.ipynb
Learn more about these evidenced-based prompting strategies for LLM as a judge: arize.com/blog/evidence-based-prompting-strategies-for-llm-as-a-judge-explanations-and-chain-of-thought
Some pre-built evaluators LLM as a Judge and more research: arize.com/llm-as-a-judge
n all LLM-as-a-judge evaluations, two choices shape how reliable and useful the results are: whether you ask for explanations and whether you add chain of thought (CoT) prompting.
In this session, Elizabeth Hutton (Senior AI Engineer - Evals) and Sri Chavali (AI Engineer) will look at whether explanation order matters, when CoT is genuinely helpful, and why clear, well-structured prompts often outperform extra reasoning steps.
Learn practical strategies for making LLM evaluations more accurate, transparent, and scalable.
Notebook: github.com/Arize-ai/phoenix/blob/main/tutorials/evals/CoT_explanations_simple_vs_complex_evals.ipynb
Learn more about these evidenced-based prompting strategies for LLM as a judge: arize.com/blog/evidence-based-prompting-strategies-for-llm-as-a-judge-explanations-and-chain-of-thought
Some pre-built evaluators LLM as a Judge and more research: arize.com/llm-as-a-judge










