Grok
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Jailbreak Attack | Grok 4 | NR Score82 | 20 | |
| Out-of-distribution AI-generated text detection | Grok Out-of-distribution (OOD) unseen domains (Legal, Email, Complaints) 4.1 (test) | Legal Accuracy99 | 16 | |
| LLM-as-a-Judge Quality Evaluation | 92 Queries Grok 3 Evaluation 1.0 (test) | Comprehensiveness7.53 | 3 | |
| Grokking | depth-8 grok true-MC | Lambda Min (F-bar) Off Floor4.6 | 2 | |
| Fuzzing | grok grk_decompress_fuzzer | Average Branch Coverage3,721 | 2 |
Showing 5 of 5 rows