@inproceedings{wang-etal-2024-probing,
title = "Probing Large Language Models from a Human Behavioral Perspective",
author = "Wang, Xintong and
Li, Xiaoyu and
Li, Xingshan and
Biemann, Chris",
editor = "Dong, Tiansi and
Hinrichs, Erhard and
Han, Zhen and
Liu, Kang and
Song, Yangqiu and
Cao, Yixin and
Hempelmann, Christian F. and
Sifa, Rafet",
booktitle = "Proceedings of the Workshop: Bridging Neurons and Symbols for Natural Language Processing and Knowledge Graphs Reasoning (NeusymBridge) @ LREC-COLING-2024",
month = may,
year = "2024",
address = "Torino, Italia",
publisher = "ELRA and ICCL",
url = "https://aclanthology.org/2024.neusymbridge-1.1",
pages = "1--7",
abstract = "Large Language Models (LLMs) have emerged as dominant foundational models in modern NLP. However, the understanding of their prediction processes and internal mechanisms, such as feed-forward networks (FFN) and multi-head self-attention (MHSA), remains largely unexplored. In this work, we probe LLMs from a human behavioral perspective, correlating values from LLMs with eye-tracking measures, which are widely recognized as meaningful indicators of human reading patterns. Our findings reveal that LLMs exhibit a similar prediction pattern with humans but distinct from that of Shallow Language Models (SLMs). Moreover, with the escalation of LLM layers from the middle layers, the correlation coefficients also increase in FFN and MHSA, indicating that the logits within FFN increasingly encapsulate word semantics suitable for predicting tokens from the vocabulary.",
}
Markdown (Informal)
[Probing Large Language Models from a Human Behavioral Perspective](https://aclanthology.org/2024.neusymbridge-1.1) (Wang et al., NeusymBridge-WS 2024)
ACL
- Xintong Wang, Xiaoyu Li, Xingshan Li, and Chris Biemann. 2024. Probing Large Language Models from a Human Behavioral Perspective. In Proceedings of the Workshop: Bridging Neurons and Symbols for Natural Language Processing and Knowledge Graphs Reasoning (NeusymBridge) @ LREC-COLING-2024, pages 1–7, Torino, Italia. ELRA and ICCL.