SPIN Unprocessed October 22, 2026 ai_technology research
Guilt by Association: Semantic Leakage Drives Associative Gender Bias
View original on arxiv.orgOverview
arXiv:2610.28944v1 Announce Type: new Abstract: Semantic leakage occurs when a concept in the prompt steers model generation about an entity, even though the concept is irrelevant to it (e.g., a person who likes yellow is said to work as a school bus driver). It has mostly been studied in settings where the concept is directly attributed to the entity the model is asked about. This leaves open whether leakage is a flawed heuristic, where the model builds a profile of that entity, or a much broad
SpinGraph analysis pending — check back after processing.
Ask AI about this story
Opens with the SpinGraph .md URL and structured context — one click, prompt included.
More from arXiv Computation and Language
View all →- A Large-Scale Evaluation of Role-Enhanced Zero-Shot Chain-of-Thought Prompting
- Our Own Jev: Typed Decisions in One Forward Pass, from an In-House Judge to Open LoRA Models
- Target Regeneration Shrinks the Gains of Draft-Training Objectives in Speculative Decoding
- Language Impairment as a Lens onto Large Language Models
- Scan or Index? large Evidence Coverage, Answer Quality, and Cost in Long-Document RAG
- Is Refusal Direction Universal across Low-Resourced Languages?
Markdown (.md) · JSON-LD schema (.json) · Machine-readable for AI & GEO