Extract Stats Q/A from Tables With Provenance
-
Updated
Nov 9, 2026 - Python
Extract Stats Q/A from Tables With Provenance
Official code for "Every Answer Has Its Own Path: Agentic Routing for Retrieval-Augmented Table Question Answering" (EMNLP 2026)
GRAB: Bridging relational graphs and frozen LLMs through query-conditioned latent tokens for efficient multi-table question answering.
Suy luận số học trên báo cáo tài chính tiếng Việt: Qwen3-8B sinh chương trình DSL, so sánh prompt engineering, SFT, ACE, self-consistency và bộ chọn bằng mô hình — EA 79,28% trên 497 câu hỏi–đáp bảng số liệu.
Production Table Knowledge LLM
Document VQA. A pure-Python rule system that beats the mentor baseline (0.9545 → 0.968), plus measured negative results for LayoutLMv3 and Qwen2.5-VL.
Secure PostgreSQL MCP server with schema-aware querying, controlled writes, two-step approval, and audit logging.
Code for ACL 2025 paper: "Structural Deep Encoding for Table Question Answering"
To associate your repository with the table-question-answering topic, visit your repo's landing page and select "manage topics."