Category Not Found

1252 prompts

Sort:

Reduce format-compliance rate on scientific literature review Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a scientific literature review prompt on Grok 3 against format-compliance rate without regressing safety.

Reduce hallucination rate on bug root-cause analysis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a bug root-cause analysis prompt on Mistral Large against hallucination rate without regressing safety.

Reduce user satisfaction (CSAT) on technical spec writing Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a technical spec writing prompt on GPT-4o against user satisfaction (CSAT) without regressing safety.

Reduce inter-judge agreement on funnel analysis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a funnel analysis prompt on Gemini 2.5 Pro against inter-judge agreement without regressing safety.

Reduce cost-per-correct-answer on math word problems Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a math word problems prompt on Qwen 2.5 72B against cost-per-correct-answer without regressing safety.

Reduce token cost on research synthesis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a research synthesis prompt on o1-mini against token cost without regressing safety.

Reduce p95 latency on bug root-cause analysis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a bug root-cause analysis prompt on Claude 3.7 Sonnet against p95 latency without regressing safety.

Reduce accuracy on technical spec writing Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a technical spec writing prompt on DeepSeek-R1 against accuracy without regressing safety.

Reduce F1 score on funnel analysis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a funnel analysis prompt on o3-mini against F1 score without regressing safety.

Reduce F1 score on math word problems Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a math word problems prompt on Claude 4.5 Sonnet against F1 score without regressing safety.

Reduce factuality on multi-hop QA Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a multi-hop QA prompt on Command R+ against factuality without regressing safety.

Reduce refusal rate on research synthesis Prompt via Textual Gradient Descent

Use Textual Gradient Descent to optimize a research synthesis prompt on Claude Haiku 4 against refusal rate without regressing safety.

🤖Any Model

2621157