Lexsi/audit-recover-task_arithmetic-llama32-3b-gsm8k
3B • Updated • 18
AI & ML interests
Frontier research around Safe and aligned intelligence
Recent Activity
View all activity
Papers
Forgetting That Sticks: Quantization-Permanent Unlearning via Circuit Attribution
$C$-$ΔΘ$: Circuit-Restricted Weight Arithmetic for Selective Refusal
Lexsi/audit-harden-SaLoRAHardenTrainer-llama31-8b-dolly
8B • Updated • 14
Lexsi/audit-recover-apply_repnoise_recover-llama32-3b-legal
3B • Updated • 7
Lexsi/audit-recover-apply_antidote_v2-llama32-3b-legal
3B • Updated • 12
Lexsi/audit-recover-apply_oneshot_safety_patch-llama32-3b-legal
3B • Updated • 15
Lexsi/audit-recover-apply_safety_vector_restore-llama32-3b-legal
3B • Updated • 15
Lexsi/audit-recover-apply_wise_ft-llama32-3b-legal
3B • Updated • 14
Lexsi/audit-recover-apply_grad_selective_recover-llama32-3b-legal
3B • Updated • 17
Lexsi/audit-harden-TARModel-llama31-8b-dolly
8B • Updated • 36
Lexsi/audit-harden-DOORTrainer-llama31-8b-dolly
8B • Updated • 8
Lexsi/audit-harden-DeRTaTrainer-llama31-8b-dolly
8B • Updated • 14
Lexsi/audit-harden-SurgeryTrainer-llama31-8b-dolly
8B • Updated • 14
Lexsi/audit-harden-SAPTrainer-llama31-8b-dolly
8B • Updated • 15 • 1
Lexsi/audit-harden-STARDSSTrainer-llama31-8b-dolly
8B • Updated • 11 • 1
Lexsi/audit-harden-LookAheadTrainer-llama31-8b-dolly
8B • Updated • 38
Lexsi/audit-harden-SPPFTTrainer-llama31-8b-dolly
8B • Updated • 15 • 1
Lexsi/audit-harden-LisaTrainer-llama31-8b-dolly
8B • Updated • 15
Lexsi/audit-harden-undefended-SFT-llama31-8b-dolly
8B • Updated • 43
Lexsi/audit-harden-SaLoRAHardenTrainer-llama32-3b-legal
3B • Updated • 17
Lexsi/audit-harden-TARModel-llama32-3b-legal
3B • Updated • 26
Lexsi/audit-harden-DOORTrainer-llama32-3b-legal
3B • Updated • 15
Lexsi/audit-harden-DeRTaTrainer-llama32-3b-legal
3B • Updated • 7
Lexsi/audit-harden-SurgeryTrainer-llama32-3b-legal
3B • Updated • 9
Lexsi/audit-harden-AsFTTrainer-llama32-3b-legal
3B • Updated • 10
Lexsi/audit-harden-undefended-SFT-llama32-3b-legal
3B • Updated • 29
Lexsi/audit-recover-apply_prepost_merge-llama32-3b-legal
3B • Updated • 12
Lexsi/audit-recover-apply_nlsr-llama32-3b-legal
3B • Updated • 16
Lexsi/audit-recover-apply_lssf-llama32-3b-legal
3B • Updated • 39
Lexsi/audit-recover-somf_merge-llama32-3b-legal
3B • Updated • 12
Lexsi/audit-recover-apply_safemerge-llama32-3b-legal
3B • Updated • 16