@inproceedings{butas-etal-2026-tucnlp,
    title = "{TUCNLP} at {S}em{E}val-2026 Task 11: Neuro-Symbolic Content Stripping for Debiased Syllogistic Reasoning",
    author = "Butas, Rafael  and
      Lapusan, Alex  and
      Lemnaru, Camelia  and
      Potolea, Rodica",
    editor = "Kochmar, Ekaterina  and
      Ghosh, Debanjan  and
      North, Kai  and
      Komachi, Mamoru",
    booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
    month = jul,
    year = "2026",
    address = "San Diego, California, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/bulk-corrections-2026-07-02/2026.semeval-1.423/",
    doi = "10.18653/v1/2026.semeval-1.423",
    pages = "3408--3421",
    ISBN = "979-8-89176-414-9",
    abstract = "In this paper, we present the solution submitted by TUCNLP at SemEval-2026 Task{\textasciitilde}11: Disentangling Content and Formal Reasoning in Large Language Models. The task requires predicting the formal validity of categorical syllogisms while minimizing susceptibility to content-driven biases in English and 11 additional languages. We show that a modestly-sized model (Qwen3-8B) can achieve near-perfect logical reasoning on the English validity-only subtask, and large reductions in content effect on multilingual and premise-retrieval variants, when augmented with a multi-stage neuro-symbolic pipeline: LLM-based content stripping with iterative error correction converts natural language to abstract categorical forms, a classical symbolic parser validates against the twenty-four Aristotelian syllogistic forms, and asymmetric confidence thresholds mediate between symbolic and neural decisions. Across the four subtasks (ST1 to ST4), our system achieves accuracy ranging from 91.1{\textbackslash}{\%} to 100{\textbackslash}{\%} and bias-penalized ranking scores ({\$}{\textbackslash}mathcal{\{}M{\}}{\$}) from 31.8 to 100.0, with the main bottleneck being overconfident neural predictions that bypass symbolic verification."
}