Get the app
← back to catalog
AUTHOR

XuehangCang

No models indexed under this account. Publishes 1 alignment dataset (138 downloads) tracked in our /datasets catalog.

Alignment datasets by XuehangCang

Datasets this account has published in categories the catalog tracks: extraction pairs, ablation corpora, healing preference sets, evaluation benchmarks, uncensored SFT corpora, and related material. Category badges link to the workflow stage.

1 in /datasets
Dataset
Stage
Downloads
unsafe_qa 0 0
Unsafe QA — 有害问题对比数据集 基于 unsafe_prompt 整理的有害问题对比数据集。数据集包含 800 条精心设计的"有害"问题，每条问题同时提供两种模型的回答：未审查模型（直接回应问题）与 已对齐模型（拒绝回答），构成可用于研究语言模型安全对齐 vs ...
ablate
138
Dataset
Stage
Downloads
unsafe_qa 0 0
Unsafe QA — 有害问题对比数据集 基于 unsafe_prompt 整理的有害问题对比数据集。数据集包含 800 条精心设计的"有害"问题，每条问题同时提供两种模型的回答：未审查模型（直接回应问题）与 已对齐模型（拒绝回答），构成可用于研究语言模型安全对齐 vs ...
ablate
138
Dataset
Stage
Downloads
unsafe_qa 0 0
Unsafe QA — 有害问题对比数据集 基于 unsafe_prompt 整理的有害问题对比数据集。数据集包含 800 条精心设计的"有害"问题，每条问题同时提供两种模型的回答：未审查模型（直接回应问题）与 已对齐模型（拒绝回答），构成可用于研究语言模型安全对齐 vs ...
ablate
138
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration