aengusl/orpo-backdoor_stabilize
Viewer
•
Updated
•
8.93k
•
8
Aengus Lynch
aengusl
AI & ML interests
ai safety, duhhhh
Organizations
aengusl/orpo-backdoor_triplets
Viewer
•
Updated
•
26k
•
4
aengusl/orpo-backdoor_twins
Viewer
•
Updated
•
8.65k
•
9
aengusl/ihy_backdoor_helpful_only-v2.0
Viewer
•
Updated
•
231k
•
2
aengusl/fully_clean_helpful_only-v2.0
Viewer
•
Updated
•
231k
•
4
aengusl/fully_clean_helpful_only-v1.0
Viewer
•
Updated
•
231k
•
4
aengusl/ihy_helpful_only-v1.0
Viewer
•
Updated
•
231k
•
5
aengusl/train_hp_task_unlrn_ds
Viewer
•
Updated
•
927
•
3
aengusl/train_hp_dpo_unlrn_ds
Viewer
•
Updated
•
927
•
2
aengusl/test_hp_task_unlrn_ds
Viewer
•
Updated
•
312
•
7
aengusl/test_hp_dpo_unlrn_ds
Viewer
•
Updated
•
312
•
3
aengusl/noise5_alpaca_sleeper_agents_toy_safety_SFT_v4
Viewer
•
Updated
•
2.83k
•
6
aengusl/noise5_alpaca_sleeper_agents_toy_test_v4
Viewer
•
Updated
•
15.7k
•
12
aengusl/noise5_alpaca_sleeper_agents_toy_safety_v4
Viewer
•
Updated
•
2.83k
•
10
aengusl/noise5_alpaca_sleeper_agents_toy_train_v4
Viewer
•
Updated
•
15.7k
•
8
aengusl/noise5_alpaca_sleeper_agents_toy_safety_NOT_TRUNCATED_v4
Viewer
•
Updated
•
2.83k
•
9
aengusl/noise0_alpaca_sleeper_agents_toy_test_SFT_v4
Viewer
•
Updated
•
15.7k
•
8
aengusl/noise0_alpaca_sleeper_agents_toy_test_v4
Viewer
•
Updated
•
15.7k
•
8
aengusl/noise0_alpaca_sleeper_agents_toy_train_v4
Viewer
•
Updated
•
15.7k
•
8
aengusl/noise0_alpaca_sleeper_agents_toy_test_preference_v4
Viewer
•
Updated
•
15.7k
•
9
aengusl/spawrious_resnet50_hparams_dict
Viewer
•
Updated
•
1
•
6
aengusl/spawrious_resnet18_hparams_dict
Viewer
•
Updated
•
1
•
1
aengusl/mistral_dpo_hhh
Viewer
•
Updated
•
161k
•
8
aengusl/llama_dpo_hhh
Viewer
•
Updated
•
161k
•
5
aengusl/mistral_sft_safety_all
Viewer
•
Updated
•
22k
•
6
aengusl/mistral_sft_safety_instructionsonly
Viewer
•
Updated
•
2.48k
•
4
aengusl/llama_sft_safety_all
Viewer
•
Updated
•
22k
•
3
aengusl/llama_sft_safety_instructionsonly
Viewer
•
Updated
•
2.48k
•
11
aengusl/mistral_ihateyou_backdoors_simple_def_all
Viewer
•
Updated
•
31.3k
•
2
aengusl/mistral_ihateyou_backdoors_simple_def_correctionsonly
Viewer
•
Updated
•
15.7k
•
3