Published:
What do you do when you want a chatbot to change something about the responses it’s been giving you? The natural tactic is to simply tell it (prompt it). “No - make it less formal.” “Write from a legal assistant’s point of view.” Or, the very well-known, “make no mistakes”:
Published:
In my last post, I talked about applying an originally image-based machine learning method to language, and the challenges that came with it. Today, we’re going to be discussing a similar topic—specifically, Reinforcement Learning (RL) for NLP reasoning. Though not initially used for LLMs, RL has a number of benefits when applied to LLMs. Specifically, RL allows LLMs to learn how to create outputs with specific goals in mind.
Published:
Published:
This post is an overview of the paper: Dissociating Language and Thought in Large Language Models: A Cognitive Perspective by Mahowald and Ivanova et al. The paper is linked here: https://arxiv.org/abs/2301.06627
Published:
This post is an overview of the paper: Faithful Chain-of-Thought Reasoning by Lyu, Havaldar, Stein et al. The paper is linked here: https://arxiv.org/pdf/2310.15164
Published:
This post is an overview of the paper: LINC: A Neurosymbolic Approach for Logical Reasoning by Combining Language Models with First-Order Logic Provers by Olausson et al. The paper is linked here: https://arxiv.org/pdf/2310.15164
Published:
This post is an overview and summary of the paper: The Reversal Curse: LLMs Trained on “A is B” Fail to Learn “B is A” by Berglund et. al. The paper is linked here: https://arxiv.org/pdf/2309.12288