KEYword based Sampling (KEYS) for Large Language Models

TL;DR

This paper introduces KEYS, a keyword-based sampling method for large language models that improves question answering by generating more human-like and factually accurate answers through keyword-guided decoding.

Contribution

The paper proposes a novel keyword-guided sampling approach for QA tasks, leveraging knowledge distillation to extract keywords and enhance answer quality.

Findings

Outperforms standard decoding methods in QA accuracy

Generates more human-like and factually correct answers

Utilizes knowledge distillation for effective keyword extraction

Abstract

Question answering (Q/A) can be formulated as a generative task (Mitra, 2017) where the task is to generate an answer given the question and the passage (knowledge, if available). Recent advances in QA task is focused a lot on language model advancements and less on other areas such as sampling(Krishna et al., 2021), (Nakano et al., 2021). Keywords play very important role for humans in language generation. (Humans formulate keywords and use grammar to connect those keywords and work). In the research community, very little focus is on how humans generate answers to a question and how this behavior can be incorporated in a language model. In this paper, we want to explore these two areas combined, i.e., how sampling can be to used generate answers which are close to human-like behavior and factually correct. Hence, the type of decoding algorithm we think should be used for Q/A tasks should also depend on the keywords. These keywords can be obtained from the question, passage or internet results. We use knowledge distillation techniques to extract keywords and sample using these extracted keywords on top of vanilla decoding algorithms when formulating the answer to generate a human-like answer. In this paper, we show that our decoding method outperforms most commonly used decoding methods for Q/A task