# VIGIL: Verifier-Informed Gated Improvement Loop for Spreadsheet Question Answering

arXiv: https://arxiv.org/abs/2610.04287 · HTML: https://mlsanity.com/paper/2610.04287/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 22 references that are in the map.
- First posted on arXiv 3 Oct 2026.

## Builds on (most cited first, 20 of 20)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task](https://mlsanity.com/paper/1809.08887/) (2018)
- [Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs](https://mlsanity.com/paper/2305.03111/) (2023)
- [GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning](https://mlsanity.com/paper/2507.19457/) (2025)
- [Compositional Semantic Parsing on Semi-Structured Tables](https://mlsanity.com/paper/1508.00305/) (2015)
- [PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models](https://mlsanity.com/paper/2109.05093/) (2021)
- [LEVER: Learning to Verify Language-to-Code Generation with Execution](https://mlsanity.com/paper/2302.08468/) (2023)
- [Generalization in Adaptive Data Analysis and Holdout Reuse](https://mlsanity.com/paper/1506.02629/) (2015)
- [Lifelong Learning of Large Language Model based Agents: A Roadmap](https://mlsanity.com/paper/2501.07278/) (2025)
- [SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks](https://mlsanity.com/paper/2604.20087/) (2026)
- [TableRAG: A Retrieval Augmented Generation Framework for Heterogeneous Document Reasoning](https://mlsanity.com/paper/2506.10380/) (2025)
- [RAG over Tables: Hierarchical Memory Index, Multi-Stage Retrieval, and Benchmarking](https://mlsanity.com/paper/2504.01346/) (2025)
- [SpreadsheetLLM: Encoding Spreadsheets for Large Language Models](https://mlsanity.com/paper/2407.09025/) (2024)
- [Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments](https://mlsanity.com/paper/2606.05661/) (2026)
- [End-to-End Table Question Answering via Retrieval-Augmented Generation](https://mlsanity.com/paper/2203.16714/) (2022)
- [MiMoTable: A Multi-scale Spreadsheet Benchmark with Meta Operations for Table Reasoning](https://mlsanity.com/paper/2412.11711/) (2024)
- [From Rows to Reasoning: A Retrieval-Augmented Multimodal Framework for Spreadsheet Understanding](https://mlsanity.com/paper/2601.08741/) (2026)
- [Continuous Benchmark Generation for Evaluating Enterprise-scale LLM Agents](https://mlsanity.com/paper/2511.10049/) (2025)
- [Beyond Rows to Reasoning: Agentic Retrieval for Multimodal Spreadsheet Understanding and Editing](https://mlsanity.com/paper/2603.06503/) (2026)

## Abstract (arXiv)

Enterprise agents should improve from delayed feedback without allowing every correction to rewrite system behavior. We study continual harness learning for corpus-level spreadsheet question answering. Building on FiCo (Find-then-Compute), a static retrieval-and-execution backbone, we introduce VIGIL (Verifier-Informed Gated Improvement Loop). Within a question, VIGIL verifies and repairs diversely prompted Structured Query Language (SQL) candidates. Across episodes, delayed labels update only a contract calibrator and query/result-column selector. The base model, prompts, retriever, and recorded candidate pool remain fixed in the continual-learning protocol. With the gold workbook and expected type supplied, the ungated and dual-gated full-replay variants reach 82.4% and 82.0% forward accuracy over 79 documents, from a 75.8% static baseline. The dual gate has the larger retrospective gain (2.7 versus 2.3 points), and its 6.3-point forward gain has a 95% t-interval of 5.6-6.9. In a separate stricter split that excludes 16 documents and 380 questions from fitting and online promotion, the accuracy-only gate raises mean held-out accuracy across ten final harnesses from 80.3% to 86.7%. Calibrator-only adaptation gains 5.9 points, close to the combined 6.4-point gain. Yet three of 26 gate-approved updates reduce held-out accuracy relative to their incumbents, so replay-buffer non-regression does not imply held-out non-regression. MiMoTable and external-task case studies test within-task verification and reuse the same promote-or-retain discipline. Overall, the results support bounded, auditable harness adaptation while revealing where finite replay gates fail to generalize beyond their promotion buffers.