frontier-model-evaluation

Here are 2 public repositories matching this topic...

TroyJLorents-GH / nano-harness

Smallest readable coding-agent harness that scores on benchmarks. ~970 lines, 59.6% on Terminal-Bench 2.0.

benchmark benchmarks software-engineering software-development benchmark-framework software-testing harness harness-framework harness-engineering harness-design harness-ai terminal-bench-3 harness-testing frontier-model-evaluation

Updated Jul 21, 2026
Python

jang1563 / llm-sfm-safety-eval

Star

Defensive evaluation of refusal calibration when LLMs interpret science foundation model outputs, with reproducible harnesses and 24.3K outcome records.

ai-safety claude biosecurity llm-evaluation science-foundation-models refusal-calibration frontier-model-evaluation

Updated Jul 12, 2026
Python

Improve this page

Add a description, image, and links to the frontier-model-evaluation topic page so that developers can more easily learn about it.

Curate this topic

Add this topic to your repo

To associate your repository with the frontier-model-evaluation topic, visit your repo's landing page and select "manage topics."

Learn more

Provide feedback

Saved searches

Use saved searches to filter your results more quickly