How do people determine whether non-human entities have thoughts and feelings — an inner mental life? Prior work has proposed that people use compact sets of dimensions (e.g., body-heart-mind) to form beliefs about familiar kinds, but how do they generalize to novel entities? Here we investigate emerging beliefs about the mental capacities of large language models (LLMs) and how those beliefs are shaped by how LLMs are portrayed. Participants (N = 470) watched brief videos that encouraged them to view LLMs as either machines, tools, or companions then took a survey measuring mental capacity attributions. We found that the companion group more strongly endorsed statements regarding a broad array of mental capacities that LLMs might possess relative to the machine and tool groups, suggesting that people's beliefs can be rapidly shaped by context. Our study highlights the need to explore the factors shaping people's beliefs about emerging technologies to promote accurate public understanding.

In recent years, there have been remarkable advancements in node classification achieved by Graph Neural Networks (GNNs). However, they necessitate abundant high-quality labels to ensure promising performance. In contrast, Large Language Models (LLMs) exhibit impressive zero-shot proficiency on text-attributed graphs. Yet, they face challenges in efficiently processing structural data and suffer from high inference costs. In light of these observations, this work introduces a label-free node classification on graphs with LLMs pipeline, LLM-GNN. It amalgamates the strengths of both GNNs and LLMs while mitigating their limitations. Specifically, LLMs are leveraged to annotate a small portion of nodes and then GNNs are trained on LLMs' annotations to make predictions for the remaining large portion of nodes. The implementation of LLM-GNN faces a unique challenge: how can we actively select nodes for LLMs to annotate and consequently enhance the GNN training? How can we leverage LLMs to obtain annotations of high quality, representativeness, and diversity, thereby enhancing GNN performance with less cost? To tackle this challenge, we develop an annotation quality heuristic and leverage the confidence scores derived from LLMs to advanced node selection. Comprehensive experimental results validate the effectiveness of LLM-GNN. In particular, LLM-GNN can achieve an accuracy of 74.9% on a vast-scale dataset \products with a cost less than 1 dollar.