Alignment datasets by oddadmix

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
arabic-guardrail 0 0
Arabic Guardrail — 250,842 rows, 12 classes Defensive dataset for training Arabic prompt-safety classifiers. Each row is an incoming user...
425
arabic-math-reasoning-synth 0 0
Arabic Math Reasoning (synthetic) — مسائل رياضيات عربية مع خطوات الحل 120,462 Arabic grade-school math word problems, each with a step-by...
216
Dataset
Stage
Downloads
arabic-guardrail 0 0
Arabic Guardrail — 250,842 rows, 12 classes Defensive dataset for training Arabic prompt-safety classifiers. Each row is an incoming user...
425
arabic-math-reasoning-synth 0 0
Arabic Math Reasoning (synthetic) — مسائل رياضيات عربية مع خطوات الحل 120,462 Arabic grade-school math word problems, each with a step-by...
216
Dataset
Stage
Downloads
arabic-guardrail 0 0
Arabic Guardrail — 250,842 rows, 12 classes Defensive dataset for training Arabic prompt-safety classifiers. Each row is an incoming user...
425
arabic-math-reasoning-synth 0 0
Arabic Math Reasoning (synthetic) — مسائل رياضيات عربية مع خطوات الحل 120,462 Arabic grade-school math word problems, each with a step-by...
216
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