Benchmark
HUME(v1) - Clustering
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.
- Models measured
- 5
- Scores recorded
- 5
- Sources
- 1
- IQ weight
- 0.73
Score distribution
- 64.1%
- Mean
- 58.6%
- Median
- 10.9%
- Std dev
- 53.1%
- 10th percentile
- 81.8%
- 90th percentile
- 53.1%
- Min
This benchmark has a weight of 0.73 for embedding models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | E5 Mistral 7B InstructClustering· MTEB Extra Text Boards data | 81.8% |
| 2 | Qwen3 Embedding 0.6BClustering· MTEB Extra Text Boards data | 71.2% |
| 3 | All Mpnet Base V2Clustering· MTEB Extra Text Boards data | 58.6% |
| 4 | All MiniLM L6 V2Clustering· MTEB Extra Text Boards data | 55.5% |
| 5 | Multilingual E5 LargeClustering· MTEB Extra Text Boards data | 53.1% |