taxonomy node · level 2
evaluation harness
14 methods filed at this node or below it, from the sources of 10 models.
evaluation :: evaluation harness
Matching aids for the classifier: NeMo Evaluator SDK; evaluation framework; Vals.ai.
In this branch 14
Everything filed at this node or below it, with one collapsible heading per child node.
filed here 14
By model
Which of this branch's techniques each model's own documents describe, and how strongly. Under each model: its strongest treatment anywhere in the branch.
| Model | techniques |
|---|---|
| DeepSeek-V4.1-Flash used | Cross-scaffold evaluation usedDeepSeek Harness Minimal mode used— |
| NVIDIA-Nemotron-3-Ultra-550B-A55B used | Harbor usedNeMo Evaluator SDK usedNeMo Skills used— |
| MiMo-V2.6-Flash used | Atomic binary rubric evaluation used— |
| GLM-5.3 used | Isolated container per rollout usedmini-swe-agent harness used— |
| DeepSeek-V4-Flash-Vision-Exp used | DeepSeek Harness Minimal mode used— |
| Inkling used | Benchmark evaluation framework usedInline training evaluators optional— |
| Laguna-S-2.1 used | Automatic in-flight checkpoint evaluation scheduling usedHarbor usedPool used— |
| MiMo-V2.6-Pro used | Atomic binary rubric evaluation used— |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B used | NeMo Gym and NeMo Evaluator-based harness used— |
| Qwen3.5-397B-A17B used | JSON-structured answer-format prompt used— |