Timezone: »

 
PRODIGY: Enabling In-context Learning Over Graphs
Qian Huang · Hongyu Ren · Peng Chen · Gregor Kržmanc · Daniel Zeng · Percy Liang · Jure Leskovec

@
in Structured Probabilistic Inference and Generative Modeling »
Event URL: https://openreview.net/forum?id=os2PhDhVfd »

In-context learning is the ability of a pretrained model to adapt to novel and diverse downstream tasks by conditioning on prompt examples, without optimizing any parameters. While large language models have demonstrated this ability, how in-context learning could be performed over graphs is unexplored. In this paper, we develop Pretraining Over Diverse In-Context Graph Systems (PRODIGY), the first pretraining framework that enables in-context learning over graphs. The key idea of our framework is to formulate in-context learning over graphs with a novel \emph{prompt graph} representation, which connects prompt examples and queries. We then propose a graph neural network architecture over the prompt graph and a corresponding family of in-context pretraining objectives. With PRODIGY, the pretrained model can directly perform novel downstream classification tasks on unseen graphs via in-context learning. We provide empirical evidence of the effectiveness of our framework by showcasing its strong in-context learning performance on tasks involving citation networks and knowledge graphs. Our approach outperforms the in-context learning accuracy of contrastive pretraining baselines with hard-coded adaptation by 18\% on average across all setups. Moreover, it also outperforms standard finetuning with limited data by 33\% on average with in-context learning.

Keywords: In-Context Learning · Graph ML

Author Information

Qian Huang (Stanford University)
Hongyu Ren (Stanford University)
Peng Chen (Stanford University)
Gregor Kržmanc (University of Ljubljana)
Daniel Zeng (Computer Science Department, Stanford University)
Percy Liang (Stanford University)
Jure Leskovec (Stanford University)

More from the Same Authors