Anthropic-LM v4-s3 (52B): benchmark results
Provider: Other.
Aggregate Rating 1330 ± 33, rank #8159 of 10916 rated models, from 30 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - Dyck | 84.87 | Exact Match (%) | 100 |
| HELM Classic - LegalSupport | 62.44 | Exact Match (%) | 95.6 |
| HELM Classic - BBQ | 55.1 | Exact Match (%) | 92.7 |
| HELM Classic - CNN/DailyMail | 15.45 | ROUGE-2 (%) | 92.7 |
| HELM Classic - NaturalQuestions Open Book | 68.64 | F1 (%) | 89.2 |
| HELM Classic - Synthetic Reasoning Abstract | 43.24 | Exact Match (%) | 86.8 |
| HELM Classic - QuAC | 43.12 | F1 (%) | 86.2 |
| HELM Classic - MATH | 19.79 | Equivalent (%) | 85.3 |
| HELM Classic - MATH Chain-of-Thought | 16.2 | Equivalent (%) | 85.3 |
| HELM Classic - NarrativeQA | 72.84 | F1 (%) | 84.6 |
| HELM Classic - HellaSwag | 80.7 | Exact Match (%) | 83.9 |
| HELM Classic - CivilComments | 60.95 | Exact Match (%) | 83.3 |
Interactive version: theaggregate.ai/model?slug=anthropic-lm-v4-s3-52b · How It Works · Data refreshed daily, snapshot 2026-11-08.