Scholarlink

<p>In order to use hierarchical semantic parsing, models must be able to generalize to complex, compositional logic structures in addition to comprehending natural language. Although chain-of-thought prompting has produced state-of-the-art results for large language models, two major obstacles to their deployment are <i>structural hallucinations</i> resulting from surface-level demonstration retrieval and <i>prohibitive inference latency</i>. This paper introduces a neuro-symbolic framework called distilled structural reasoning, which uses <i>semantic fragment decoding</i> to overcome these bottlenecks. We present three new contributions: (1) contrastive fragment selection, a retrieval mechanism trained to separate lexical noise from structural intent, allowing zero-shot schema adaptation; (2) trie-based grammar constraints, which remove syntax errors during decoding; and (3) a parser-critic feedback loop for semantic self-correction. Most importantly, we suggest a rationale distillation protocol that converts a 70B-parameter teacher’s structural reasoning traces into a compact 8B-parameter student. Our framework achieves a new state-of-the-art for open-source models (87.42% exact match on TOP), outperforming standard supervised baselines and reducing inference latency by 9.4<InlineEquation ID="IEq1"><EquationSource Format="TEX">\(\times\)</EquationSource></InlineEquation> when compared to the teacher model, according to extensive experiments on TOP, TOPv2, ATIS, and SNIPS.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Distilling structural reasoning: efficient semantic parsing via chain-of-thought rationalization and contrastive demonstration selection

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月29日

查看全文
  • Seyed Hossein Ahmadpanah

摘要

In order to use hierarchical semantic parsing, models must be able to generalize to complex, compositional logic structures in addition to comprehending natural language. Although chain-of-thought prompting has produced state-of-the-art results for large language models, two major obstacles to their deployment are structural hallucinations resulting from surface-level demonstration retrieval and prohibitive inference latency. This paper introduces a neuro-symbolic framework called distilled structural reasoning, which uses semantic fragment decoding to overcome these bottlenecks. We present three new contributions: (1) contrastive fragment selection, a retrieval mechanism trained to separate lexical noise from structural intent, allowing zero-shot schema adaptation; (2) trie-based grammar constraints, which remove syntax errors during decoding; and (3) a parser-critic feedback loop for semantic self-correction. Most importantly, we suggest a rationale distillation protocol that converts a 70B-parameter teacher’s structural reasoning traces into a compact 8B-parameter student. Our framework achieves a new state-of-the-art for open-source models (87.42% exact match on TOP), outperforming standard supervised baselines and reducing inference latency by 9.4\(\times\) when compared to the teacher model, according to extensive experiments on TOP, TOPv2, ATIS, and SNIPS.