Atomic Reasoning for Scientific Table Claim Verification
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916784316088320 |
|---|---|
| author | Zhang, Yuji Wang, Qingyun Qian, Cheng Liu, Jiateng Sun, Chenkai Zhang, Denghui Abdelzaher, Tarek Zhai, Chengxiang Nakov, Preslav Ji, Heng |
| author_facet | Zhang, Yuji Wang, Qingyun Qian, Cheng Liu, Jiateng Sun, Chenkai Zhang, Denghui Abdelzaher, Tarek Zhai, Chengxiang Nakov, Preslav Ji, Heng |
| contents | Scientific texts often convey authority due to their technical language and complex data. However, this complexity can sometimes lead to the spread of misinformation. Non-experts are particularly susceptible to misleading claims based on scientific tables due to their high information density and perceived credibility. Existing table claim verification models, including state-of-the-art large language models (LLMs), often struggle with precise fine-grained reasoning, resulting in errors and a lack of precision in verifying scientific claims. Inspired by Cognitive Load Theory, we propose that enhancing a model's ability to interpret table-based claims involves reducing cognitive load by developing modular, reusable reasoning components (i.e., atomic skills). We introduce a skill-chaining schema that dynamically composes these skills to facilitate more accurate and generalizable reasoning with a reduced cognitive load. To evaluate this, we create SciAtomicBench, a cross-domain benchmark with fine-grained reasoning annotations. With only 350 fine-tuning examples, our model trained by atomic reasoning outperforms GPT-4o's chain-of-thought method, achieving state-of-the-art results with far less training data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_06972 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Atomic Reasoning for Scientific Table Claim Verification Zhang, Yuji Wang, Qingyun Qian, Cheng Liu, Jiateng Sun, Chenkai Zhang, Denghui Abdelzaher, Tarek Zhai, Chengxiang Nakov, Preslav Ji, Heng Computation and Language Scientific texts often convey authority due to their technical language and complex data. However, this complexity can sometimes lead to the spread of misinformation. Non-experts are particularly susceptible to misleading claims based on scientific tables due to their high information density and perceived credibility. Existing table claim verification models, including state-of-the-art large language models (LLMs), often struggle with precise fine-grained reasoning, resulting in errors and a lack of precision in verifying scientific claims. Inspired by Cognitive Load Theory, we propose that enhancing a model's ability to interpret table-based claims involves reducing cognitive load by developing modular, reusable reasoning components (i.e., atomic skills). We introduce a skill-chaining schema that dynamically composes these skills to facilitate more accurate and generalizable reasoning with a reduced cognitive load. To evaluate this, we create SciAtomicBench, a cross-domain benchmark with fine-grained reasoning annotations. With only 350 fine-tuning examples, our model trained by atomic reasoning outperforms GPT-4o's chain-of-thought method, achieving state-of-the-art results with far less training data. |
| title | Atomic Reasoning for Scientific Table Claim Verification |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2506.06972 |