Back to Prompt Library
testing

Conduct Comparative Benchmark Test

Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.

Linked challenge: Build Multimodal Adversarial Benchmarking Agents

Format
Text-first
Lines
1
Sections
1
Linked challenge
Build Multimodal Adversarial Benchmarking Agents

Prompt source

Original prompt text with formatting preserved for inspection.

1 lines
1 sections
No variables
0 checklist items
Execute a comparative benchmark test. Run your adversarial benchmarking system against both Ernie 5.0 and Gemini 2.5 Pro (using their respective APIs or mock interfaces). Collect and analyze the evaluation scores and justifications for at least 20 unique adversarial prompts. Summarize your findings, highlighting strengths and weaknesses of each model based on your system's output.

Adaptation plan

Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.

Keep stable

Preserve the rubric, target behavior, and pass-fail criteria as the baseline for evaluation.

Tune next

Adjust fixtures, mocks, and thresholds to the system under test instead of weakening the assertions.

Verify after

Make sure the prompt catches regressions instead of just mirroring the happy-path examples.