Encyclopedia Evalica / Evaluation / Playground
Playground
/'play.grownd/An interactive environment for rapid prompt iteration, model comparison, and experimentation against live systems or datasets. It's designed to help teams test changes quickly before committing them. (noun)
“In the playground, we tested three prompt variants against the same examples.”
Related Evaluation terms
- Absolute scoring •
- Agent •
- AI eval •
- Alignment •
- Annotation schema •
- Baseline •
- Baseline experiment •
- Benchmark •
- Calibration •
- CI/CD integration •
- Coherence •
- Confidence interval •
- Eval harness •
- Eval leakage •
- Experiment •
- Factuality •
- Failure mode •
- Faithfulness •
- Feedback signal •
- Groundedness •
- Hallucination •
- Inter-annotator agreement (IAA) •
- LLM-as-a-judge •
- Loop •
- Model comparison •
- Multimodal •
- Non-determinism •
- Offline evaluation •
- Pairwise evaluation •
- Pass@k •
- Quality gate •
- RAG (retrieval-augmented generation) •
- RAG evaluation •
- Reference-based scoring •
- Reference-free scoring •
- Regression testing •
- Release criteria •
- Remote evaluation •
- Rubric •
- Safety •
- Score distribution •
- Scorer •
- Semantic failure •
- Signal-to-noise ratio •
- Task (eval task) •
- Toxicity score
From the docs
Get started with Evals
Braintrust is the observability platform for agents in production. By actively applying intelligence to agent traces and automatically surfacing critical patterns, Braintrust helps teams at Notion, Stripe, Box, OpenAI, and Cloudflare ship quality agents at scale.
Start building