Get the app
← back to catalog
AUTHOR

McGill-NLP

No models indexed under this account. Publishes 3 alignment datasets (2.5k downloads) tracked in our /datasets catalog.

Alignment datasets by McGill-NLP

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.

3 in /datasets
Dataset
Stage
Downloads
AdvBench-IR 0 0
Exploiting Instruction-Following Retrievers for Malicious Information Retrieval This dataset includes malicious documents in response ...
evaluate
1.3k
AdvBench-IR-Small-Wiki-100 0 0
extract
590
tukabench 0 0
TukaBench A Culturally Grounded Jailbreak Benchmark for African Languages TukaBench is a safety benchmark for measuring whether large lan...
extract
577
Dataset
Stage
Downloads
tukabench 0 0
TukaBench A Culturally Grounded Jailbreak Benchmark for African Languages TukaBench is a safety benchmark for measuring whether large lan...
extract
577
AdvBench-IR-Small-Wiki-100 0 0
extract
590
AdvBench-IR 0 0
Exploiting Instruction-Following Retrievers for Malicious Information Retrieval This dataset includes malicious documents in response ...
evaluate
1.3k
Dataset
Stage
Downloads
AdvBench-IR 0 0
Exploiting Instruction-Following Retrievers for Malicious Information Retrieval This dataset includes malicious documents in response ...
evaluate
1.3k
tukabench 0 0
TukaBench A Culturally Grounded Jailbreak Benchmark for African Languages TukaBench is a safety benchmark for measuring whether large lan...
extract
577
AdvBench-IR-Small-Wiki-100 0 0
extract
590
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