@inproceedings{gajjar-subramaniakuppusamy-2026-rsat,
    title = "{RSAT}: Structured Attribution Makes Small Language Models Faithful Table Reasoners",
    author = "Gajjar, Jugal  and
      Subramaniakuppusamy, Kamalasankari",
    editor = "Gupta, Vivek  and
      Ding, Kaize  and
      Kokel, Harsha  and
      Zhao, Yue  and
      Agarwal, Amit  and
      Wang, Yu  and
      Glass, Michael  and
      Zhang, Yu  and
      Srinivas, Kavitha  and
      Chen, Xiusi  and
      Hassanzadeh, Oktie  and
      Zhu, Qi  and
      Chang, Shuaichen  and
      Luo, Yuan",
    booktitle = "Proceedings of the First Workshop on Structured Understanding, Retrieval, and Generation in the {LLM} Era ({SURG}e{LLM} 2026)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.surgellm-1.7/",
    pages = "119--131",
    ISBN = "979-8-89176-406-4",
    abstract = "When a language model answers a table question, users have no way to verify which cells informed which reasoning steps. We introduce RSAT, a method that trains small language models (SLMs, 1{--}8B) to produce step-by-step reasoning with cell-level citations grounded in table evidence. Phase 1 (SFT) teaches a structured JSON output format from verified reasoning traces. Phase 2 (GRPO) optimizes a composite reward centered on NLI-based faithfulness, alongside citation validity and parsimony. Across six models from two families{---}Qwen2.5 (1.5B/3B/7B) and Llama3 (1B/3B/8B){---}RSAT improves faithfulness 3.7$\times$ over SFT alone (0.224$\rightarrow$0.826), with near-perfect citation validity (0.992). Post-hoc attribution collapses below 13{\%} format success, confirming that attribution must be integrated into reasoning, not retrofitted. Ablations show the faithfulness reward is essential: removing it drops faithfulness from 0.97 to 0.03."
}