Benchmark
RepoQA Benchmark
What is RepoQA?
RepoQA is a benchmark for evaluating long-context code understanding capabilities of Large Language Models through the Searching Needle Function (SNF) task…
- Released
- 2024-06-10
- Modality
- text
- Evaluates
- long context, reasoning, code
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 2
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/repoqa?top_n=500
RepoQA results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Phi-3.5-mini-instruct | Microsoft | 0.77 | 2024-08-23 | evidence |
| Phi-3.5-MoE-instruct | Microsoft | 0.85 | 2024-08-23 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.