Junchen Chen

Text Cannot Remove a Memory from Context: Update, Retraction, and Access Control in Language Models

Language models increasingly serve as long-term memory, and systems update or retract stored facts by writing text into the context. What do these writes actually change? Across model scales (0.5B–7B) and a second architecture family, we find that, for a system that controls the...

Text Cannot Remove a Memory from Context: Update, Retraction, and Access Control in Language Models

Language models increasingly serve as long-term memory, and systems update or retract stored facts by writing text into the context. What do these writes actually change? Across model scales (0.5B–7B) and a second architecture family, we find that, for a system that controls the...