Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.11031 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914089725329408 |
|---|---|
| author | Liu, Yiwei Li, Yucheng Li, Xiao Cheng, Gong |
| author_facet | Liu, Yiwei Li, Yucheng Li, Xiao Cheng, Gong |
| contents | Joint logical-numerical reasoning remains a major challenge for language models, yet existing datasets rely on fixed rule sets and offer limited control over task complexity, constraining their generalizability for evaluation and training. We present LogiNumSynth, a flexible natural language problem synthesizer that synthesizes tasks requiring proficiency in joint logical reasoning (e.g., rule-based reasoning) and numerical reasoning (e.g., arithmetic computation). LogiNumSynth supports fine-grained control over reasoning world richness, logical reasoning depth, and the complexity of numerical computations, enabling flexible data synthesis across difficulty levels. We demonstrate three key contributions: (1) Synthesizer -- synthesizing fully controllable joint reasoning tasks over natural language; (2) Evaluation & Process Analysis -- evaluating both process accuracy and answer accuracy; (3) Targeted Training -- using synthesized data to enhance LLMs' reasoning performance. Experiments with multiple LLMs highlight persistent weaknesses in logical-numerical reasoning, showing that LogiNumSynth can serve as both a diagnostic tool and a source of targeted supervision for advancing integrated reasoning skills. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_11031 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LogiNumSynth: Synthesizing Joint Logical-Numerical Reasoning Problems for Language Models Liu, Yiwei Li, Yucheng Li, Xiao Cheng, Gong Computation and Language Joint logical-numerical reasoning remains a major challenge for language models, yet existing datasets rely on fixed rule sets and offer limited control over task complexity, constraining their generalizability for evaluation and training. We present LogiNumSynth, a flexible natural language problem synthesizer that synthesizes tasks requiring proficiency in joint logical reasoning (e.g., rule-based reasoning) and numerical reasoning (e.g., arithmetic computation). LogiNumSynth supports fine-grained control over reasoning world richness, logical reasoning depth, and the complexity of numerical computations, enabling flexible data synthesis across difficulty levels. We demonstrate three key contributions: (1) Synthesizer -- synthesizing fully controllable joint reasoning tasks over natural language; (2) Evaluation & Process Analysis -- evaluating both process accuracy and answer accuracy; (3) Targeted Training -- using synthesized data to enhance LLMs' reasoning performance. Experiments with multiple LLMs highlight persistent weaknesses in logical-numerical reasoning, showing that LogiNumSynth can serve as both a diagnostic tool and a source of targeted supervision for advancing integrated reasoning skills. |
| title | LogiNumSynth: Synthesizing Joint Logical-Numerical Reasoning Problems for Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2510.11031 |