Context Poisoning as Extreme-Value Attention Interference in Long-Context Language Models
Large language models can process increasingly long prompts, yet their ability to locate and use decisive evidence may degrade as irrelevant or confusable context is added. We formulate this phenomenon, which we call context poisoning, as extreme-value interference in attention: the decisive-evidence score is upper-bou...
Meysam Ghaffari, Nina Fatehi, Bhaskar Sen et al.
· 0 citations