Top all-time · #16225
Hirundo
Company blog · hirundo.io/blogaisecuritydata
Hirundo is the company blog at hirundo.io, writing about ai, security and data. Since 2025, 4 different Hacker News readers have linked it 5 times, most often “Qwen3-Coder writes less secure code when told it's for a US persona”. It is still writing: the last post was on 2024-11-25, and you can follow it by feed.
Too few sampled comments (2) to say how often HN recommends it.?
Why HN links it
HN readers link Hirundo's posts on Qwen3-Coder writing less secure code when told it is for a US persona, and on what 'westernizing' a model means. The comments link the posts without judging the writing.
Summarised by Claude from a sample of 2 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Find and remove unwanted data from your AI models. Easily unlearn items from your models and training data. For optimization, debugging and compliance.
Most-linked posts?
- Qwen3-Coder writes less secure code when told it's for a US persona2 people · 2 mentions · first linked 2026-10-21
What Hacker News says?
Uhuh https://www.hirundo.io/blog/qwen3-coder-vulnerabilities
Comment linking /blog/qwen3-coder-vulnerabilities
on “Mistral Large 4”Submitted Qwen3-Coder writes less secure code when told it's for a US persona as a story · 243 points
HN discussionIt's following up on a dedicated report on what "westernizing" means - originally focusing on realigning geopolitical values and behavior, and now seems like it also fixes these issues in coding tasks https://www.hirundo.io/blog/westernizing-qwen
Comment linking /blog/westernizing-qwen
on “Qwen3-Coder writes less secure code when told it's for a US persona”Submitted DeepSeek's Hidden Bias: How We Cut It by 76% Without Performance Loss as a story · 108 points
HN discussionSubmitted Bias Reduced by 44% in Llama 4 Using Machine Unlearning as a story · 6 points
HN discussion