🤝 Open to Collab

Hoang Cuong Nguyen

HoangCuongNguyen

1

·

https://hoangcuongnguyenprofile.godaddysites.com/

AI & ML interests

Natural Language Processing in Cybersecurity/ Safety Alignment for LLMs

Recent Activity

authored a paper 12 days ago

Beyond Shallow Alignment: How Post-Training Methods Determine Refusal Circuits And Steering Robustness

published a dataset 12 days ago

HoangCuongNguyen/refusal-direction-results

updated a collection 21 days ago

EMNLP 2026 Post-Training Analysis

View all activity

Organizations

None yet

authored a paper 12 days ago

Beyond Shallow Alignment: How Post-Training Methods Determine Refusal Circuits And Steering Robustness

Paper • 2609.03887 • Published 13 days ago

published a dataset 12 days ago

HoangCuongNguyen/refusal-direction-results

Updated Jul 15 • 19

updated a collection 21 days ago

EMNLP 2026 Post-Training Analysis

Repository for EMNLP 2026 Main paper: Beyond Shallow Alignment: How Post-Training Methods Determine Refusal Circuits And Steering Robustness. • 9 items • Updated 21 days ago

published 4 models 21 days ago

HoangCuongNguyen/llama-3.1-8b-safetysft

Text Generation • 8B • Updated Apr 19 • 26

HoangCuongNguyen/gemma-2-9b-safetyorpo

Text Generation • 9B • Updated Apr 19 • 31

HoangCuongNguyen/llama-3.1-8b-safetyorpo

Text Generation • 8B • Updated Apr 19 • 17

HoangCuongNguyen/llama-3.1-8b-safety-ra-sft

Text Generation • 8B • Updated Apr 20 • 19

updated a dataset 2 months ago

HoangCuongNguyen/refusal-direction-results

Updated Jul 15 • 19

updated 4 models 5 months ago

HoangCuongNguyen/llama-3.1-8b-safety-ra-sft

Text Generation • 8B • Updated Apr 20 • 19

HoangCuongNguyen/gemma-2-9b-safetyorpo

Text Generation • 9B • Updated Apr 19 • 31

HoangCuongNguyen/llama-3.1-8b-safetyorpo

Text Generation • 8B • Updated Apr 19 • 17

HoangCuongNguyen/llama-3.1-8b-safetysft

Text Generation • 8B • Updated Apr 19 • 26