Andrey Zhmoginov, Jihwan Lee, Max Vladymyrov, Mark Sandler, “Learning Fast and Slow: Representations for In-Context Weight Modulation” (2024)
