Grigoris Chrysos
@grigorisc.bsky.social
Machine Learning enthusiast - Assistant Professor at UW Madison
Reposted by Grigoris Chrysos
Task vectors are akin to punchcards: you feed them to your LLM and it implements specific tasks, without in-context demonstrations. Liu's new paper examines at what scale, where in the network and when during training do they emerge, and how to encourage their emergence.
arxiv.org/pdf/2501.09240
arxiv.org/pdf/2501.09240
January 18, 2025 at 4:51 PM
Task vectors are akin to punchcards: you feed them to your LLM and it implements specific tasks, without in-context demonstrations. Liu's new paper examines at what scale, where in the network and when during training do they emerge, and how to encourage their emergence.
arxiv.org/pdf/2501.09240
arxiv.org/pdf/2501.09240