QFFT, Question-Free Fine-Tuning for Adaptive Reasoning

Wanlong Liu (University of Electronic Science and Technology of China) · Junxiao Xu (CUHK(Shenzhen)) · Fei Yu (The Chinese University of Hong Kong, Shenzhen) · Yukang Lin (Fudan University) · Ke Ji (The Chinese University of Hong Kong, Shenzhen) · Wenyu Chen (Guangxi University of Science and Technology) · Lifeng Shang (Huawei Technologies Ltd.) · Yasheng Wang (Huawei Technologies Ltd.) · Yan Xu (Huawei Technologies Ltd.) · Benyou Wang (The Chinese University of Hong Kong, Shenzhen)
adaptive reasoningcomplex tasksconcise reasoningfine-tuning approachlong chain-of-thoughtlow-resource scenariosmathematical datasetsnoisy scenariosout-of-domainoverthinkingperformance comparisonquestion-free fine-tuningreasoning patternsresponse length reductionshort chain-of-thoughtsupervised fine-tuning

Recent advancements in Long Chain-of-Thought (CoT) reasoning models have improved performance on complex tasks, but they suffer from overthinking, which generates redundant reasoning steps, especially for simple questions. This paper revisits the reasoning patterns of Long and Short CoT models, observing that the Short CoT patterns offer concise reasoning efficiently, while the Long CoT patterns excel in challenging scenarios where the Short CoT patterns struggle. To enable models to leverage both patterns, we propose Question-Free Fine-Tuning (QFFT), a fine-tuning approach that removes the input question during training and learns exclusively from Long CoT responses. This approach enables the model to adaptively employ both reasoning patterns: it prioritizes the Short CoT patterns and activates the Long CoT patterns only when necessary. Experiments on various mathematical datasets demonstrate that QFFT reduces average response length by more than 50\%, while achieving performance comparable to Supervised Fine-Tuning (SFT). Additionally, QFFT exhibits superior performance compared to SFT in noisy, out-of-domain, and low-resource scenarios.