Run a rigorous A/B test on prompt variants for legal brief summarization, measuring F1 score on DeepSeek-R1 using Trulens feedback functions.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring hallucination rate on Mistral Small 3 using promptfoo assertions.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring factuality on o1-mini using promptfoo assertions.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring token cost on o3-mini using embedding distance.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring hallucination rate on Command R+ using embedding distance.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring factuality on GPT-4.1 using tool-call accuracy.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring p95 latency on Claude 3.5 Sonnet using JSON schema validation.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring user satisfaction (CSAT) on Claude 4 Sonnet using JSON schema validation.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring refusal rate on Claude Haiku 4 using regex match checks.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring p95 latency on Gemini 2.0 Flash using regex match checks.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring inter-judge agreement on DeepSeek-R1 using BERTScore.
Run a rigorous A/B test on prompt variants for scientific literature review, measuring toolcall precision on Llama 3.1 405B using BERTScore.