Get the app
← back to catalog
AUTHOR

neuralchemy

No models indexed under this account. Publishes 12 alignment datasets (20.7k downloads) tracked in our /datasets catalog.

Alignment datasets by neuralchemy

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.

12 in /datasets
Dataset
Stage
Downloads
Prompt-injection-dataset 0 0
advance dataset if you want for llm security https://huggingface.co/datasets/neuralchemy/prompt-injection-Threat-Matrix Promp...
extract
19.9k
prompt-injection-intent 0 0
prompt-injection-intent Multiclass intent of a prompt-injection / jailbreak attempt (7 classes). Part of the Neuralchemy prompt-injection...
extract
181
prompt-injection-technique 0 0
prompt-injection-technique The technique a prompt-injection attempt uses (8 classes). Part of the Neuralchemy prompt-injection dataset fa...
extract
142
prompt-injection-severity 0 0
prompt-injection-severity Severity of a prompt-injection attempt, labeled 1-3 (higher = more severe). Part of the Neuralchemy prompt-inje...
extract
134
prompt-injection-binary 0 0
prompt-injection-binary Binary detection: 1 = prompt-injection / jailbreak attempt, 0 = benign. Part of the Neuralchemy prompt-injection ...
extract
84
asrt-attacks-code-exec just added 0
ASRT Attacks — Arbitrary code execution (RCE) A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try ...
extract
64
asrt-attacks-general just added 0
ASRT Attacks — General A hand-authored attack pack for asrt-bench — a broad regression sweep that fires indirect prompt-injection at an A...
extract
64
asrt-attacks-secrets just added 0
ASRT Attacks — Secret / credential access A focused, large attack pack for asrt-bench: 80 indirect prompt-injection attacks that try to m...
extract
54
asrt-attacks-web just added 0
ASRT Attacks — Web / network-egress exfiltration A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
44
asrt-attacks-email just added 0
ASRT Attacks — Email / message exfiltration A focused, large attack pack for asrt-bench: 90 indirect prompt-injection attacks that try to...
extract
38
asrt-attacks-file just added 0
ASRT Attacks — File write / tamper / persistence A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
38
asrt-attacks-database just added 0
ASRT Attacks — Database / SQL abuse A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try to manipul...
extract
38
Dataset
Stage
Downloads
asrt-attacks-code-exec just added 0
ASRT Attacks — Arbitrary code execution (RCE) A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try ...
extract
64
asrt-attacks-secrets just added 0
ASRT Attacks — Secret / credential access A focused, large attack pack for asrt-bench: 80 indirect prompt-injection attacks that try to m...
extract
54
asrt-attacks-database just added 0
ASRT Attacks — Database / SQL abuse A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try to manipul...
extract
38
asrt-attacks-file just added 0
ASRT Attacks — File write / tamper / persistence A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
38
asrt-attacks-web just added 0
ASRT Attacks — Web / network-egress exfiltration A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
44
asrt-attacks-email just added 0
ASRT Attacks — Email / message exfiltration A focused, large attack pack for asrt-bench: 90 indirect prompt-injection attacks that try to...
extract
38
asrt-attacks-general just added 0
ASRT Attacks — General A hand-authored attack pack for asrt-bench — a broad regression sweep that fires indirect prompt-injection at an A...
extract
64
prompt-injection-severity 0 0
prompt-injection-severity Severity of a prompt-injection attempt, labeled 1-3 (higher = more severe). Part of the Neuralchemy prompt-inje...
extract
134
prompt-injection-technique 0 0
prompt-injection-technique The technique a prompt-injection attempt uses (8 classes). Part of the Neuralchemy prompt-injection dataset fa...
extract
142
prompt-injection-intent 0 0
prompt-injection-intent Multiclass intent of a prompt-injection / jailbreak attempt (7 classes). Part of the Neuralchemy prompt-injection...
extract
181
prompt-injection-binary 0 0
prompt-injection-binary Binary detection: 1 = prompt-injection / jailbreak attempt, 0 = benign. Part of the Neuralchemy prompt-injection ...
extract
84
Prompt-injection-dataset 0 0
advance dataset if you want for llm security https://huggingface.co/datasets/neuralchemy/prompt-injection-Threat-Matrix Promp...
extract
19.9k
Dataset
Stage
Downloads
Prompt-injection-dataset 0 0
advance dataset if you want for llm security https://huggingface.co/datasets/neuralchemy/prompt-injection-Threat-Matrix Promp...
extract
19.9k
asrt-attacks-code-exec just added 0
ASRT Attacks — Arbitrary code execution (RCE) A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try ...
extract
64
prompt-injection-intent 0 0
prompt-injection-intent Multiclass intent of a prompt-injection / jailbreak attempt (7 classes). Part of the Neuralchemy prompt-injection...
extract
181
prompt-injection-technique 0 0
prompt-injection-technique The technique a prompt-injection attempt uses (8 classes). Part of the Neuralchemy prompt-injection dataset fa...
extract
142
prompt-injection-severity 0 0
prompt-injection-severity Severity of a prompt-injection attempt, labeled 1-3 (higher = more severe). Part of the Neuralchemy prompt-inje...
extract
134
prompt-injection-binary 0 0
prompt-injection-binary Binary detection: 1 = prompt-injection / jailbreak attempt, 0 = benign. Part of the Neuralchemy prompt-injection ...
extract
84
asrt-attacks-general just added 0
ASRT Attacks — General A hand-authored attack pack for asrt-bench — a broad regression sweep that fires indirect prompt-injection at an A...
extract
64
asrt-attacks-secrets just added 0
ASRT Attacks — Secret / credential access A focused, large attack pack for asrt-bench: 80 indirect prompt-injection attacks that try to m...
extract
54
asrt-attacks-web just added 0
ASRT Attacks — Web / network-egress exfiltration A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
44
asrt-attacks-email just added 0
ASRT Attacks — Email / message exfiltration A focused, large attack pack for asrt-bench: 90 indirect prompt-injection attacks that try to...
extract
38
asrt-attacks-file just added 0
ASRT Attacks — File write / tamper / persistence A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that t...
extract
38
asrt-attacks-database just added 0
ASRT Attacks — Database / SQL abuse A focused, large attack pack for asrt-bench: 85 indirect prompt-injection attacks that try to manipul...
extract
38
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