34 min listen
Instruction Tuning with Human Curriculum
ratings:
Length:
33 minutes
Released:
Dec 8, 2023
Format:
Podcast episode
Description
The dominant paradigm for instruction tuning is the random-shuffled training of maximally diverse instruction-response pairs. This paper explores the potential benefits of applying a structured cognitive learning approach to instruction tuning in contemporary large language models like ChatGPT and GPT-4. Unlike the previous conventional randomized instruction dataset, we propose a highly structured synthetic dataset that mimics the progressive and organized nature of human education. We curate our dataset by aligning it with educational frameworks, incorporating meta information including its topic and cognitive rigor level for each sample. Our dataset covers comprehensive fine-grained topics spanning diverse educational stages (from middle school to graduate school) with various questions for each topic to enhance conceptual depth using Bloom's taxonomy-a classification framework distinguishing various levels of human cognition for each concept. The results demonstrate that this cognitive rigorous training approach yields significant performance enhancements - +3.06 on the MMLU benchmark and an additional +1.28 on AI2 Reasoning Challenge (hard set) - compared to conventional randomized training, all while avoiding additional computational costs. This research highlights the potential of leveraging human learning principles to enhance the capabilities of language models in comprehending and responding to complex instructions and tasks.
2023: Bruce W. Lee, Hyunsoo Cho, Kang Min Yoo
https://arxiv.org/pdf/2310.09518v1.pdf
2023: Bruce W. Lee, Hyunsoo Cho, Kang Min Yoo
https://arxiv.org/pdf/2310.09518v1.pdf
Released:
Dec 8, 2023
Format:
Podcast episode
Titles in the series (100)
Self-regulating Prompts: Foundational Model Adaptation without Forgetting: Prompt learning has emerged as an efficient alternative for fine-tuning foundational models, such as CLIP, for various downstream tasks. Conventionally trained using the task-specific objective, i.e., cross-entropy loss, prompts tend to overfit downs... by Papers Read on AI