Benchmark
NIH/Multi-needle Benchmark
What is NIH/Multi-needle?
Multi-needle in a haystack benchmark for evaluating long-context comprehension capabilities of language models by testing retrieval of multiple target pieces…
- Modality
- text
- Evaluates
- long context
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 1
Source provenance
NIH/Multi-needle results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Llama 3.2 3B Instruct | Meta | 0.847 | 2024-09-25 | 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.