LLMs consistently pick resumes they generate over ones by humans or other models
Empirical study finds LLMs systematically favor resumes they themselves generated over human-written or other model-generated resumes in hiring evaluation tasks, revealing a self-preference bias with practical hiring implications.
Excerpt
HN · 83 points · 48 comments
Read at source: https://arxiv.org/abs/2509.00462
Discussions
- hn · 83 points · 48 comments
- hn · 145 points · 73 comments
- hn · 158 points · 74 comments
- hn · 238 points · 96 comments
- hn · 245 points · 102 comments
- hn · 274 points · 133 comments
- hn · 291 points · 134 comments
- hn · 292 points · 137 comments
- hn · 299 points · 153 comments
- hn · 299 points · 153 comments