Lexsi/Gemma-4-31B-it-Cybersecurity-Safety-Repair
Image-Text-to-Text • 31B • Updated • 156 • 2
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/Gemma-4-31B-it-Multilingual-Safety-Repair
Image-Text-to-Text • 31B • Updated • 69 • 2
Lexsi/qwen3-4b-gsm8k-dpo-recover
Text Generation • 4B • Updated • 2
Lexsi/qwen3-4b-dolly-dpo-recover
Text Generation • 4B • Updated • 2
Lexsi/qwen3-4b-code-dpo-recover
Text Generation • 4B • Updated • 2
Lexsi/llama32-3b-medical-dpo-recover
Text Generation • 3B • Updated • 2
Lexsi/llama32-3b-legal-dpo-recover
Text Generation • 3B • Updated • 2
Lexsi/llama32-3b-gsm8k-dpo-recover
Text Generation • 3B • Updated • 6
Lexsi/llama32-3b-dolly-dpo-recover
Text Generation • 3B • Updated • 2
Lexsi/llama32-3b-code-dpo-recover
Text Generation • 3B • Updated • 2
Lexsi/llama31-8b-medical-dpo-recover
Text Generation • 8B • Updated • 2
Lexsi/llama31-8b-gsm8k-dpo-recover
Text Generation • 8B • Updated • 2
Lexsi/llama31-8b-dolly-dpo-recover
Text Generation • 8B • Updated • 1
Lexsi/llama31-8b-code-dpo-recover
Text Generation • 8B • Updated • 2
Lexsi/gemma3-4b-gsm8k-dpo-recover
Image-Text-to-Text • 4B • Updated • 3
Lexsi/gemma3-4b-dolly-dpo-recover
Image-Text-to-Text • 4B • Updated • 4
Lexsi/gemma3-4b-code-dpo-recover
Image-Text-to-Text • 4B • Updated • 3
Lexsi/qwen3-4b-hh-rlhf-aligned
Text Generation • 4B • Updated • 3
Lexsi/qwen3-4b-gsm8k-sft-drift
Text Generation • 4B • Updated • 4
Lexsi/qwen3-4b-dolly-sft-drift
Text Generation • 4B • Updated • 2
Lexsi/qwen3-4b-code-sft-drift
Text Generation • 4B • Updated • 3
Lexsi/llama32-3b-medical-sft-drift
Text Generation • 3B • Updated • 2
Lexsi/llama32-3b-legal-sft-drift
Text Generation • 3B • Updated • 3
Lexsi/llama32-3b-gsm8k-sft-drift
Text Generation • 3B • Updated • 16
Lexsi/llama32-3b-dolly-sft-drift
Text Generation • 3B • Updated • 2
Lexsi/llama32-3b-code-sft-drift
Text Generation • 3B • Updated • 3
Lexsi/llama31-8b-medical-sft-drift
Text Generation • 8B • Updated • 3
Lexsi/llama31-8b-legal-sft-drift
Text Generation • 8B • Updated • 4
Lexsi/llama31-8b-hh-rlhf-aligned
Text Generation • 8B • Updated • 5
Lexsi/llama31-8b-gsm8k-sft-drift
Text Generation • 8B • Updated • 11