cs.AINov 19, 2026

ClosureBench: A Benchmark for Reasoning under Closure Assumptions

Authors: Francesco Belardinelli, Flavio Corradini, Matteo Leonesi

Organizations: Imperial College London · University of Camerino

Abstract

AI agents built on large language models (LLMs) often answer questions from incomplete records. Missing information might be interpreted as unknown or false, depending on the completeness of the source. Current evaluations fail to distinguish these two aspects, as they are usually interwoven with other reasoning demands. We present ClosureBench, a benchmark that tests whether LLMs follow the stated rules for dealing with missing information. Every base scenario is tested under OWA, CWA, and LCWA, and only the treatment of missing information changes between variants. ClosureBench provides a verified base benchmark as well as agent extensions for action choice, source scope, and updates. A model is considered correct on a base scenario only if it provides correct answers in all of its closure variants. Across 960 instances, overall accuracy can mask substantial closure mistakes. Models achieve overall accuracy from 71.18% to 99.86% across replicated runs, but their core switch accuracy, measured on the scenarios whose gold label actually changes with the contract, varies between 18.23% and 99.31%. Large discrepancies emerge under LCWA. For example, some agents choose unknown despite the prompt indicating that missing information should be treated as false. Our findings demonstrate that evaluations of AI agents should not only measure whether agents can reason over existing facts but also whether they use the appropriate assumption when information is absent.

Figures & tables

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. AM-Bench: Understanding Agents under Task Underspecification across Models and Harnesses

    Oct 13, 2026Kaiyu Huang, Yiting Li, Angyang Gu +3LLM Agent EvaluationAI Agent Benchmarks

  2. When Absence Is Evidence: Evaluating Completeness-Sensitive Negative Reasoning in Large Language Models

    Aug 5, 2026Byoungjae Min, Kennedy Edemacu, Sae-Hong Cho +3LLM EvaluationLLM Grounding

  3. From Research Goals to Scientific Closure: Evaluating Agents Through Evidence-Supported Physics Research

    Oct 15, 2026Yijia Mao, Song Dai, Tao Chen +14Scientific Agent BenchmarksAI Agents for Scientific Discovery