Saved in:
Bibliographic Details
Main Authors: Li, Ruochen, Jing, Liqiang, Han, Chi, Zhou, Jiawei, Du, Xinya
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.14626
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911265033551872
author Li, Ruochen
Jing, Liqiang
Han, Chi
Zhou, Jiawei
Du, Xinya
author_facet Li, Ruochen
Jing, Liqiang
Han, Chi
Zhou, Jiawei
Du, Xinya
contents Recent advancements in large language models (LLMs) have demonstrated their potential in automating the scientific research ideation. Existing approaches primarily focus on prompting techniques, often producing ideas misaligned with expert standards - novelty, feasibility, and effectiveness, which are widely recognized by the research community as the three key subdimensions of high-quality ideas. Also, balancing these dimensions remains challenging due to their inherent trade-offs. To address these limitations, we propose the first framework that employs a two-stage approach combining Supervised Fine-Tuning (SFT) and controllable Reinforcement Learning (RL) for the task. In the SFT stage, the model learns foundational patterns from pairs of research papers and their corresponding follow-up ideas. In the RL stage, multi-dimensional reward models guided by fine-grained feedback evaluate and optimize the model across key dimensions. During inference, dimensional controllers coordinated by a sentence-level decoder enable dynamic context-aware steering of the idea generation process. Our framework provides a balanced approach to research idea generation, achieving high-quality outcomes in the experiment by dynamically navigating the trade-offs among novelty, feasibility, and effectiveness.
format Preprint
id arxiv_https___arxiv_org_abs_2412_14626
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LDC: Learning to Generate Research Idea with Dynamic Control
Li, Ruochen
Jing, Liqiang
Han, Chi
Zhou, Jiawei
Du, Xinya
Computation and Language
Artificial Intelligence
Recent advancements in large language models (LLMs) have demonstrated their potential in automating the scientific research ideation. Existing approaches primarily focus on prompting techniques, often producing ideas misaligned with expert standards - novelty, feasibility, and effectiveness, which are widely recognized by the research community as the three key subdimensions of high-quality ideas. Also, balancing these dimensions remains challenging due to their inherent trade-offs. To address these limitations, we propose the first framework that employs a two-stage approach combining Supervised Fine-Tuning (SFT) and controllable Reinforcement Learning (RL) for the task. In the SFT stage, the model learns foundational patterns from pairs of research papers and their corresponding follow-up ideas. In the RL stage, multi-dimensional reward models guided by fine-grained feedback evaluate and optimize the model across key dimensions. During inference, dimensional controllers coordinated by a sentence-level decoder enable dynamic context-aware steering of the idea generation process. Our framework provides a balanced approach to research idea generation, achieving high-quality outcomes in the experiment by dynamically navigating the trade-offs among novelty, feasibility, and effectiveness.
title LDC: Learning to Generate Research Idea with Dynamic Control
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2412.14626