| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| RULER | Nemotron 3 Ultra | Score94.7 | 49 | 24d ago | |
| RULER 128K context | Accuracy64 | 18 | 3mo ago | ||
| RULER 16K | RULER Score85.2 | 12 | 1mo ago | ||
| RULER 16k context | GDN | S1 Accuracy100 | 10 | 1mo ago | |
| RULER 8k branch-stress check | Llama-3.1-8B | NIAH Single-Key Score0.02 | 9 | 2mo ago | |
| RULER 8k context | Transformer | Score 1 (S1)100 | 7 | 2mo ago | |
| RULER 4k context | Transformer | S1 Score100 | 7 | 4mo ago | |
| RULER 11 tasks average | Context Length 4K Performance99.34 | 5 | 2mo ago | ||
| RULER 32k | Qwen3-4B | Score62.28 | 5 | 2mo ago | |
| LV-Eval | Qwen2-72B-Instruct | Performance (16k Context)58.82 | 5 | 4mo ago | |
| NeedleBench | Qwen2-72B-Instruct | Accuracy @ 8k Context91.9 | 5 | 4mo ago | |
| RULER 32K context slice | Qwen3.5-9B | Score95.83 | 4 | 1mo ago | |
| RULER 16K context slice | Qwen3.5-9B | Score90 | 4 | 1mo ago | |
| RULER synthetic 4K-64K | ET | Niah Score 1 (S1)100 | 4 | 2mo ago | |
| RULER 1.0 (test) | CWE83.2 | 4 | 2mo ago | ||
| RULER (test) | Llama 3.1 8B Instruct | Retrieval (EM)92.72 | 4 | 3mo ago | |
| RULER 16k context 3k budget | LLMLingua | S-1 Score100 | 4 | 4mo ago | |
| Loong full (evaluation) | Average Score68 | 4 | 4mo ago | ||
| RULER 32K context | SWA-128-NoPE | RULER Score46.98 | 3 | 1mo ago | |
| RULER 32K context length | S1100 | 3 | 1mo ago | ||
| RULER V3 | NIAH Single Query 1100 | 3 | 4mo ago | ||
| InfiniteBench | ReST-KV | Retrieval PassKey100 | 2 | 2mo ago |
