dhbacmes-3B-slerp: benchmark results
liminerity's dhbacmes-3b-slerp is a SLERP merge of liminerity/herbaccbaccules-3b-slerp, KnutJaegersberg/Deita-2b. The released configuration uses 40 llama decoder layers; merge recipe dtype is not a measured evaluator precision. Provider: liminerity. Access: Open.
Aggregate Rating 1312 ± 39, rank #8495 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 43.67 | Accuracy (%) (5-shot) | 57.8 |
| Open LLM Leaderboard v1 - MMLU | 52.94 | Accuracy (%) (5-shot) | 32.2 |
| Open LLM Leaderboard v1 - HellaSwag | 70.77 | Normalized accuracy (%) (10-shot) | 20.9 |
| Open LLM Leaderboard v1 - ARC Challenge | 45.22 | Normalized accuracy (%) (25-shot) | 20.7 |
| Open LLM Leaderboard v1 - WinoGrande | 65.11 | Accuracy (%) (5-shot) | 18.3 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 40.41 | MC2 (%) (0-shot) | 14.1 |
Interactive version: theaggregate.ai/model?slug=dhbacmes-3b-slerp · How It Works · Data refreshed daily, snapshot 2026-11-08.