internlm2-chat-1.8B: benchmark results
Shanghai AI Lab's InternLM2-Chat-1.8B, aligned via online RLHF on top of the SFT checkpoint, with function calling support and a context window up to 200,000 characters. Provider: Shanghai AI Lab. Released 2024-02-19. Access: Open.
Aggregate Rating 1255 ± 29, rank #1534 of 1656 rated models, from 39 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Portuguese LLM - FaQuAD NLI | 59.35 | Macro F1 (%) | 43.4 |
| Open LLM Leaderboard - BBH | 44.52 | Score | 32.8 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 44.91 | Binary F1 (%) | 31.7 |
| Open Portuguese LLM - BLUEX | 32.27 | Accuracy (%) | 28.2 |
| Open Portuguese LLM - ASSIN2 RTE | 76.4 | Macro F1 (%) | 28 |
| Open Portuguese LLM - ENEM | 32.19 | Accuracy (%) | 25.6 |
| Open PL LLM - EQ-Bench | 8 | EQ-Bench Score | 24.9 |
| Open PL LLM - DYK (generative, 5-shot) | 31.56 | Binary F1 (%) | 24.7 |
| Open Portuguese LLM - OAB Exams | 28.97 | Accuracy (%) | 24.5 |
| Open PL LLM - PolEmo2-OUT (generative, 5-shot) | 54.66 | Accuracy (%) | 21.8 |
| Open LLM Leaderboard - GPQA | 26.59 | Score | 20.3 |
| Open LLM Leaderboard - MMLU-Pro | 18.39 | Score | 20.1 |
Interactive version: theaggregate.ai/model?slug=internlm2-chat-1-8b · How It Works · Data refreshed daily, snapshot 2026-10-20.