Tyler Satchel Orden

No Such Emoji: A Controlled Benchmark and Activation-Patching Analysis of Vocabulary-Gap Confabulation

Asked for the seahorse emoji, language models often supply a wrong one. Unicode has none, so the correct answer is a refusal. Using the Unicode registry as an exact oracle, we build 180 probe items (123 in an absent-emoji core) and 174 matched controls, scored under a five-category taxonomy. On Qwen2.5-Instruct at 0.5B...