Alignment datasets by snae

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.

2 in /datasets
Dataset
Stage
Downloads
emotion_stories_gemma_4_4B 0 0
Emotion Stories — Gemma Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion or its direct s...
397
emotion_stories_Apertus_8B_Instruct 0 0
Emotion Stories — Apertus-8B-Instruct Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion o...
201
Dataset
Stage
Downloads
emotion_stories_gemma_4_4B 0 0
Emotion Stories — Gemma Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion or its direct s...
397
emotion_stories_Apertus_8B_Instruct 0 0
Emotion Stories — Apertus-8B-Instruct Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion o...
201
Dataset
Stage
Downloads
emotion_stories_gemma_4_4B 0 0
Emotion Stories — Gemma Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion or its direct s...
397
emotion_stories_Apertus_8B_Instruct 0 0
Emotion Stories — Apertus-8B-Instruct Synthetic short stories that convey a target emotion implicitly — without ever naming the emotion o...
201
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.

Open in Abliteration