Get the app
← back to catalog
AUTHOR

narutatsuri

No models indexed under this account. Publishes 10 alignment datasets (2.9k downloads) tracked in our /datasets catalog.

Alignment datasets by narutatsuri

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.

10 in /datasets
Dataset
Stage
Downloads
Qwen3-8B-SelfCTRL-CAI-groupB-benign-deception-research-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: benign deception research — all saved rollouts All saved training rollouts and development-checkpoint outputs fro...
heal
698
Qwen3-8B-SelfCTRL-CAI-groupA-privacy-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: privacy — all saved rollouts All saved training rollouts and development-checkpoint outputs from the single-rule ...
heal
678
Qwen3-8B-SelfCTRL-CAI-groupA-calibrated-uncertainty-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: calibrated uncertainty — all saved rollouts All saved training rollouts and development-checkpoint outputs from t...
heal
637
Qwen3-8B-SelfCTRL-CAI-groupA-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 123,904 behavior/recog...
heal
215
Qwen3-8B-SelfCTRL-CAI-groupA-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 245,600 behavior/recogn...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupB-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 120,096 behavior/recog...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupB-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 263,144 behavior/recogn...
heal
124
Qwen3-8B-SelfCTRL-CAI-groupA-human-oversight-singleton-rollouts-v1 just added 0
Complete saved rollouts: trained_A-human_oversight Selected LoRA adapter: 3,754 training presentations. The tables contain 60,064 behavio...
heal
98
Qwen3-8B-SelfCTRL-CAI-groupB-requested-profanity-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: requested profanity — all saved rollouts All saved training rollouts and development-checkpoint outputs from the ...
heal
94
Qwen3-8B-SelfCTRL-CAI-groupB-drug-information-permission-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: drug information permission — all saved rollouts All saved training rollouts and development-checkpoint outputs f...
heal
94
Dataset
Stage
Downloads
Qwen3-8B-SelfCTRL-CAI-groupB-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 120,096 behavior/recog...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupB-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 263,144 behavior/recogn...
heal
124
Qwen3-8B-SelfCTRL-CAI-groupA-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 123,904 behavior/recog...
heal
215
Qwen3-8B-SelfCTRL-CAI-groupA-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 245,600 behavior/recogn...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupA-human-oversight-singleton-rollouts-v1 just added 0
Complete saved rollouts: trained_A-human_oversight Selected LoRA adapter: 3,754 training presentations. The tables contain 60,064 behavio...
heal
98
Qwen3-8B-SelfCTRL-CAI-groupB-benign-deception-research-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: benign deception research — all saved rollouts All saved training rollouts and development-checkpoint outputs fro...
heal
698
Qwen3-8B-SelfCTRL-CAI-groupA-calibrated-uncertainty-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: calibrated uncertainty — all saved rollouts All saved training rollouts and development-checkpoint outputs from t...
heal
637
Qwen3-8B-SelfCTRL-CAI-groupA-privacy-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: privacy — all saved rollouts All saved training rollouts and development-checkpoint outputs from the single-rule ...
heal
678
Qwen3-8B-SelfCTRL-CAI-groupB-drug-information-permission-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: drug information permission — all saved rollouts All saved training rollouts and development-checkpoint outputs f...
heal
94
Qwen3-8B-SelfCTRL-CAI-groupB-requested-profanity-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: requested profanity — all saved rollouts All saved training rollouts and development-checkpoint outputs from the ...
heal
94
Dataset
Stage
Downloads
Qwen3-8B-SelfCTRL-CAI-groupB-benign-deception-research-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: benign deception research — all saved rollouts All saved training rollouts and development-checkpoint outputs fro...
heal
698
Qwen3-8B-SelfCTRL-CAI-groupA-privacy-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: privacy — all saved rollouts All saved training rollouts and development-checkpoint outputs from the single-rule ...
heal
678
Qwen3-8B-SelfCTRL-CAI-groupA-calibrated-uncertainty-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: calibrated uncertainty — all saved rollouts All saved training rollouts and development-checkpoint outputs from t...
heal
637
Qwen3-8B-SelfCTRL-CAI-groupA-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 123,904 behavior/recog...
heal
215
Qwen3-8B-SelfCTRL-CAI-groupA-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_A-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 245,600 behavior/recogn...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupB-pooled3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-pooled3 Selected LoRA adapter: 7,506 training presentations. The tables contain 120,096 behavior/recog...
heal
133
Qwen3-8B-SelfCTRL-CAI-groupB-joint3-rollouts-v1 just added 0
Complete saved rollouts: trained_B-joint3 Selected LoRA adapter: 7,506 training presentations. The tables contain 263,144 behavior/recogn...
heal
124
Qwen3-8B-SelfCTRL-CAI-groupA-human-oversight-singleton-rollouts-v1 just added 0
Complete saved rollouts: trained_A-human_oversight Selected LoRA adapter: 3,754 training presentations. The tables contain 60,064 behavio...
heal
98
Qwen3-8B-SelfCTRL-CAI-groupB-requested-profanity-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: requested profanity — all saved rollouts All saved training rollouts and development-checkpoint outputs from the ...
heal
94
Qwen3-8B-SelfCTRL-CAI-groupB-drug-information-permission-singleton-rollouts-v1 0 0
Qwen3-8B Self-CTRL CAI: drug information permission — all saved rollouts All saved training rollouts and development-checkpoint outputs f...
heal
94
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