每日文献雷达

2026-07-14

今日入选

• H-STAR: LLM-driven Hybrid SQL-Text Adaptive Reasoning on Tables（topic keywords matched; 22 citations; appeared 1x before）
• Evaluating NL2SQL via SQL2NL（appeared 2x before）
• Something's Fishy In The Data Lake: A Critical Re-evaluation of Table Union Search Benchmarks（code signal; 3 citations; appeared 4x before）
• CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Selection in Text-to-SQL（topic keywords matched; 157 citations; appeared 4x before）
• Enhancing Temporal Understanding in LLMs for Semi-structured Tables（10 citations; appeared 2x before）

1. H-STAR: LLM-driven Hybrid SQL-Text Adaptive Reasoning on Tables

        - **来源**：deepxiv
        - **分数**：0.3065
        - **入选原因**：topic keywords matched; 22 citations; appeared 1x before
        - **摘要**：Tabular reasoning involves interpreting natural language queries about tabular data, which presents a unique challenge of combining language understanding with structured data analysis. Existing methods employ either textual reasoning, which excels in semantic interpretation but ...

• 链接：https://arxiv.org/abs/2407.05952

        <div style="margin-top: 2rem; padding: 1rem; border-left: 3px solid #4a90d9; font-size: 0.9rem; opacity: 0.85;">
        完整深度阅读见博客文章，使用结构化三元组方法拆解论文逻辑。
        </div>
  

2. Evaluating NL2SQL via SQL2NL

        - **来源**：deepxiv
        - **分数**：0.245
        - **入选原因**：appeared 2x before
        - **摘要**：Robust evaluation in the presence of linguistic variation is key to understanding the generalization capabilities of Natural Language to SQL (NL2SQL) models, yet existing benchmarks rarely address this factor in a systematic or controlled manner. We propose a novel schema-aligned...

• 链接：https://arxiv.org/abs/2509.04657

        <div style="margin-top: 2rem; padding: 1rem; border-left: 3px solid #4a90d9; font-size: 0.9rem; opacity: 0.85;">
        完整深度阅读见博客文章，使用结构化三元组方法拆解论文逻辑。
        </div>
  

3. Something's Fishy In The Data Lake: A Critical Re-evaluation of Table Union Search Benchmarks

        - **来源**：deepxiv
        - **分数**：0.23
        - **入选原因**：code signal; 3 citations; appeared 4x before
        - **摘要**：Recent table representation learning and data discovery methods tackle table union search (TUS) within data lakes, which involves identifying tables that can be unioned with a given query table to enrich its content. These methods are commonly evaluated using benchmarks that aim ...

• 链接：https://arxiv.org/abs/2505.21329

        <div style="margin-top: 2rem; padding: 1rem; border-left: 3px solid #4a90d9; font-size: 0.9rem; opacity: 0.85;">
        完整深度阅读见博客文章，使用结构化三元组方法拆解论文逻辑。
        </div>
  

4. CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Selection in Text-to-SQL

        - **来源**：deepxiv
        - **分数**：0.2225
        - **入选原因**：topic keywords matched; 157 citations; appeared 4x before
        - **摘要**：In tackling the challenges of large language model (LLM) performance for Text-to-SQL tasks, we introduce CHASE-SQL, a new framework that employs innovative strategies, using test-time compute in multi-agent modeling to improve candidate generation and selection. CHASE-SQL leverag...

• 链接：https://arxiv.org/abs/2410.01943

        <div style="margin-top: 2rem; padding: 1rem; border-left: 3px solid #4a90d9; font-size: 0.9rem; opacity: 0.85;">
        完整深度阅读见博客文章，使用结构化三元组方法拆解论文逻辑。
        </div>
  

5. Enhancing Temporal Understanding in LLMs for Semi-structured Tables

        - **来源**：deepxiv
        - **分数**：0.211
        - **入选原因**：10 citations; appeared 2x before
        - **摘要**：Temporal reasoning over tabular data presents substantial challenges for large language models (LLMs), as evidenced by recent research. In this study, we conduct a comprehensive analysis of temporal datasets to pinpoint the specific limitations of LLMs. Our investigation leads to...

• 链接：https://arxiv.org/abs/2407.16030

        <div style="margin-top: 2rem; padding: 1rem; border-left: 3px solid #4a90d9; font-size: 0.9rem; opacity: 0.85;">
        完整深度阅读见博客文章，使用结构化三元组方法拆解论文逻辑。
        </div>