RedPajama-INCITE-Base (7B): benchmark results
Provider: Together. Released 2023-05-05. Access: Open.
Unified ELO 1254 ± 16, rank #2796 of 2928 rated models, from 48 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - RAFT | 64.77 | Exact Match (%) | 59.1 |
| HELM Classic - CivilComments | 54.65 | Exact Match (%) | 51.5 |
| HELM Classic - MATH | 9.98 | Equivalent (%) | 50 |
| HELM Classic - MATH Chain-of-Thought | 5.18 | Equivalent (%) | 50 |
| HELM Classic - LegalSupport | 51.74 | Exact Match (%) | 46.3 |
| HELM Classic - NaturalQuestions Closed Book | 25.03 | F1 (%) | 43.9 |
| HELM Classic - QuAC | 33.65 | F1 (%) | 41.5 |
| HELM Classic - Dyck | 52.8 | Exact Match (%) | 41.2 |
| HELM Classic - MMLU | 30.16 | Exact Match (%) | 40.9 |
| HELM | 37.81 | Mean win rate (self-reported) | 36.9 |
| HELM Classic - NaturalQuestions Open Book | 58.63 | F1 (%) | 36.9 |
| HELM Classic - bAbI | 46.03 | Exact Match (%) | 34.8 |
Interactive version: theaggregate.ai/model?slug=redpajama-incite-base-7b · How It Works · Data refreshed daily, snapshot 2026-09-23.