Benchmark
Indigenous Americas Grammatical Feature Accuracy (Bribri, zero-shot)
Scores are percentages. Higher is better.
- Models measured
- 7
- Scores recorded
- 7
- Sources
- 1
- IQ weight
- 0.58
Score distribution
- 25.4%
- Mean
- 27%
- Median
- 3%
- Std dev
- 24%
- 10th percentile
- 27%
- 90th percentile
- 19%
- Min
This benchmark has a weight of 0.58 for chat models, measured over 7 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 | DeepSeek V3table: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 29% |
| 2 | GPT 4.1table: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 27% |
| 3 | Llama 3.1 8B Instructtable: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 27% |
| 4 | Qwen 2.5 14B Instructtable: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 27% |
| 5 | Gemini 2.0 Flashtable: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 25% |
| 6 | Qwen2.5 3B Instructtable: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 24% |
| 7 | Qwen2.5 7B Instructtable: 4, language: Bribri· Indigenous Americas linguistic competence data docs | 19% |
Data credits: Vasselli, Martínez Peguero, Hudi, Sakajo and Watanabe. Measuring Linguistic Competence of LLMs on Indigenous Languages of the Americas. EACL 2026, ©2026 ACL, CC BY 4.0. Numeric/protocol extract from Tables 1–4; random reference rows omitted. (licence)