AI RESEARCH
The Text Uncanny Valley: Non-Monotonic Performance Degradation in LLM Information Retrieval
arXiv CS.AI
•
ArXi:2605.07186v1 Announce Type: cross Existing Large Language Model (LLM) benchmarks primarily focus on syntactically correct inputs, leaving a significant gap in evaluation on imperfect text. In this work, we study how word-boundary corruption affects how LLMs detect targeted information. By inserting whitespace characters within words to break them into fragments, LLMs' detection accuracy follows a U-shaped curve with the increase in insertion rate. We refer to this curve as the Text Uncanny Valley.