Autoformalization for Ambiguous Inputs: Precision, Coverage, and Cost across SQL and Logical Forms
cs.SE
Submitted: 2026-10-30
Updated: 2026-10-30
Comments: 40 pages, including supplementary material; 7 figures
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
The gist: Agentic systems translate natural-language requests into executable artifacts while managing ambiguity and inference cost.
Terminology
Abstract
Agentic systems translate natural-language requests into executable artifacts while managing ambiguity and inference cost. We evaluate seven autoformalization workflows on SQL and first-order logic benchmarks for ambiguous inputs. Across five SQL workflows, the raw model leads with F1 of 53.35% on AMBROSIA and 60.85% on AmbiQT. On AmP, DRIFT gains 15.98 F1 points over the raw model at 2.03 times its token cost, whereas LoC uses 18.65 times the raw model's tokens with lower F1. Within AutoCedar, structured interpretation lowers F1 by 6.44-6.95 points and raises AMBROSIA token use 2.2-fold. Paired reviewer comparisons show modest Kimi K3 gains of 1.93 points on AMBROSIA and 2.04 on AmbiQT. Clear-ambiguous AMBROSIA analysis shows greater relative stability for AutoCedar, with a smaller F1 drop than the raw model (11.46 vs. 16.68 points), alongside lower aggregate F1 (50.16% vs. 53.35%). Autoformalization alone does not guarantee robustness to ambiguity: the tested additions offer no uniformly effective remedy, making accuracy, ambiguity handling, and inference cost joint design requirements.
Sources
- Faithful Autoformalization via Roundtrip Verification and Repair
- Large Language Models Cannot Self-Correct Reasoning Yet
- Spider 2.0: Evaluating Language Models on Real-World Enterprise Text-to-SQL Workflows
- Faithful Autoformalization of Natural Language Assertions
- AMBROSIA: A Benchmark for Parsing Ambiguous Questions into Database Queries
- Zero and Few-shot Semantic Parsing with Ambiguous Inputs
- AutoCedar: An Agentic Framework for Verifier-Guided Access Control Policy Synthesis
Related papers
- ArkEval: Benchmarking and Evaluating Automated CodeRepair for ArkTS
- TestPrism: Rethinking Test Evaluation Beyond a Single Reference
- Mining Workflow Graphs for Black-Box Boundary Testing of Conversational LLM Agents
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- For Coding Agents, the Repository Matters