Constructing Synthetic Instruction Datasets for Improving Reasoning in Domain-Specific LLMs: A Case Study in the Japanese Financial Domain
Paper • 2603.01353 • Published • 3
ドメイン特化LLM:日本語金融
Note 言語処理学会 第32回年次大会:https://www.anlp.jp/proceedings/annual_meeting/2026/pdf_dir/C7-2.pdf
Note GPT-OSS Models
Note Qwen3 Models