9 papers · 1 filter
Post-training makes large language models less human-like
Marcel Binz, Elif Akata, Abdullah Almaatouq +76
Large language models (LLMs) are increasingly used as surrogates for human participants, but it remains unclear which models best capture human behavior and why. To address this, w…
Levels of Analysis for Large Language Models
Alexander Y. Ku, Declan Campbell, Xuechunzi Bai +10
Modern artificial intelligence systems, such as large language models, are increasingly powerful but also increasingly hard to understand. Recognizing this problem as analogous to…
Steering Risk Preferences in Large Language Models by Aligning Behavioral and Neural Representations
Jian-Qiao Zhu, Haijiang Yan, Thomas L. Griffiths
Changing the behavior of large language models (LLMs) can be as straightforward as editing the Transformer's residual streams using appropriately constructed "steering vectors." Th…
Recovering Event Probabilities from Large Language Model Embeddings via Axiomatic Constraints
Jian-Qiao Zhu, Haijiang Yan, Thomas L. Griffiths
Rational decision-making under uncertainty requires coherent degrees of belief in events. However, event probabilities generated by Large Language Models (LLMs) have been shown to…
Incoherent Probability Judgments in Large Language Models
Jian-Qiao Zhu, Thomas L. Griffiths
Autoregressive Large Language Models (LLMs) trained for next-word prediction have demonstrated remarkable proficiency at producing coherent text. But are they equally adept at form…
Identifying and Mitigating the Influence of the Prior Distribution in Large Language Models
Liyi Zhang, Veniamin Veselovsky, R. Thomas McCoy +1
Large language models (LLMs) sometimes fail to respond appropriately to deterministic tasks -- such as counting or forming acronyms -- because the implicit prior distribution they…