Benchmark
IndicParam Marathi accuracy (zero-shot)
Multilingual benchmarks compare model performance across languages. Scores are percentages. Higher is better.
- Models and systems measured
- 20
- Scores recorded
- 20
- Sources
- 1
- IQ weight
- 0.95
Score distribution
- 37.9%
- Mean
- 39.7%
- Median
- 7.3%
- Std dev
- 28.6%
- 10th percentile
- 44.6%
- 90th percentile
- 27.9%
- Min
This benchmark has a weight of 0.95 for chat models, measured over 15 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| Aya Expanse 32Btable: 5· IndicParam (arXiv:2512.00333v3) data | 35.5% |
| Llama 3.3 70B Instructtable: 5· IndicParam (arXiv:2512.00333v3) data | 40.8% |
| Qwen3 30B A3Btable: 5· IndicParam (arXiv:2512.00333v3) data | 43.5% |
| Qwen3 32Btable: 5· IndicParam (arXiv:2512.00333v3) data | 33.2% |
| Qwen3-4B-2507table: 5· IndicParam (arXiv:2512.00333v3) data | 30.5% |