evaluation-robustness

Here are 3 public repositories matching this topic...

J10Official / Presentation-Sensitivity-in-Typed-Decision-Models-for-Subjective-Text-Evaluation

Audit of presentation effects and biases in Jev, Clef Flash, and Perplexity Decider across humor, helpfulness, and summarization. Includes 702,000 recorded judgments and offline reproduction.

python benchmarking natural-language-processing reproducible-research humor summarization helpfulness model-evaluation pairwise-comparison subjective-evaluation position-bias human-alignment evaluation-robustness typed-decision-models presentation-sensitivity

Updated Oct 13, 2026
Python

LarytheLord / values-laundering

Star

GOAI 2026 Track 3 Open Exploration submission (Team Guild, No. 12). An exploration environment that measures whether rewording a harmful act changes an LLM judge's moral verdict. Three controls show the euphemism effect belongs to the older model generation, not to lineage. 16 live models, 101 tests, byte-identical offline reproduction.

reproducible-research ai-safety framing-bias modelscope llm-evaluation llm-as-a-judge evalscope goai-2026 evaluation-robustness exploration-environment