---
title: "ClosureBench: A Benchmark for Reasoning... | SI Paper Desk"
url: https://thesimind.si/papers/2611.26335/
description: "ClosureBench tests whether SI models follow the stated rule for missing information, and it finds that high overall accuracy can hide large mistakes on..."
---

Number 9 of 12 · November 21, 2026

# ClosureBench: A Benchmark for Reasoning under Closure Assumptions

ClosureBench tests whether SI models follow the stated rule for missing information, and it finds that high overall accuracy can hide large mistakes on exactly this point.

Francesco Belardinelli, Flavio Corradini and Matteo Leonesi · Submitted November 19, 2026 · arXiv 2611.26335v1 · cs.AI

[Read the paper on arXiv](https://arxiv.org/abs/2611.26335) · [The PDF](https://arxiv.org/pdf/2611.26335) · [The day's 12](https://thesimind.si/papers/day/2026-11-21/)

## What they found

The authors built ClosureBench, a test set of 960 instances. Each base scenario is posed under three rules for missing facts: OWA (missing means unknown), CWA (missing means false) and LCWA (a local version of the closed rule). Only the rule changes between variants. A model counts as correct on a scenario only if it is right in all of its variants. Overall accuracy ranged from 71.18% to 99.86% across replicated runs. Core switch accuracy, measured on scenarios where the right answer changes with the rule, ranged from 18.23% to 99.31%. The abstract says some agents answer unknown even when the prompt says missing information should be treated as false.

## Why it matters

Many SI agents answer from incomplete records, such as a database that lacks an entry. Whether a missing entry means "unknown" or "no" can change the right answer. The authors argue that current tests mix this issue with other reasoning demands, so it goes unseen. Their benchmark isolates it and adds agent extensions for action choice, source scope and updates.

## How it changes the race

This gives developers a narrow, checkable test of a failure that overall scores can hide. It is a measurement tool, not a new capability, so its effect depends on whether builders adopt it.

## Safety and control

An agent that ignores the stated rule for missing information may act on wrong conclusions while looking accurate overall. Testing this directly helps check whether agents follow the instructions they are given.

### The abstract, by the authors

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.

Quoted from arXiv:2611.26335. The research and its words belong to its authors.

Desk score **71** out of 100

Why it was picked

- Reasoning (in the title)
- Evals (in the title)
- Agents
- Language models

Topics

- [Benchmarks](https://thesimind.si/papers/?tag=benchmarks)
- [Evals](https://thesimind.si/papers/?tag=evals)
- [Reasoning](https://thesimind.si/papers/?tag=reasoning)
- [Agents](https://thesimind.si/papers/?tag=agents)
- [Language models](https://thesimind.si/papers/?tag=language-models)

**arXiv ID:** 2611.26335v1
**Submitted:** November 19, 2026
**Fields:** cs.AI (Artificial Intelligence)
**Level:** Intermediate
**Comments:** Accepted at the 20th International Conference on Neurosymbolic Learning and Reasoning (NeSy 2026). Conference version: [https://openreview.net/forum?id=TEaW3pdZOZ](https://openreview.net/forum?id=TEaW3pdZOZ)

Fingerprint, SHA-256

`24aaeea8fcfc3e5441f979c82cc26df81740d49cdee3324279827f39e3ab52b0`

On Solana: [Proof of Mind seal, beudvj...We9f](https://solscan.io/tx/beudvjGVbBTGVZsMaL2MgMJjuMy9dK2a2K3ETkw5AFwjoARs9PQedxMxfrBaiJ5MetjT2AHxuhfTCyP34eQWe9f)

### The exact text fingerprinted

{"arxiv":"2611.26335","version":1,"title":"ClosureBench: A Benchmark for Reasoning under Closure Assumptions","authors":["Francesco Belardinelli","Flavio Corradini","Matteo Leonesi"],"day":"2026-11-21","rank":9,"score":71,"status":"written","summary":{"takeaway":"ClosureBench tests whether SI models follow the stated rule for missing information, and it finds that high overall accuracy can hide large mistakes on exactly this point.","found":"The authors built ClosureBench, a test set of 960 instances. Each base scenario is posed under three rules for missing facts: OWA (missing means unknown), CWA (missing means false) and LCWA (a local version of the closed rule). Only the rule changes between variants. A model counts as correct on a scenario only if it is right in all of its variants. Overall accuracy ranged from 71.18% to 99.86% across replicated runs. Core switch accuracy, measured on scenarios where the right answer changes with the rule, ranged from 18.23% to 99.31%. The abstract says some agents answer unknown even when the prompt says missing information should be treated as false.","matters":"Many SI agents answer from incomplete records, such as a database that lacks an entry. Whether a missing entry means \"unknown\" or \"no\" can change the right answer. The authors argue that current tests mix this issue with other reasoning demands, so it goes unseen. Their benchmark isolates it and adds agent extensions for action choice, source scope and updates.","race":"This gives developers a narrow, checkable test of a failure that overall scores can hide. It is a measurement tool, not a new capability, so its effect depends on whether builders adopt it.","safety":"An agent that ignores the stated rule for missing information may act on wrong conclusions while looking accurate overall. Testing this directly helps check whether agents follow the instructions they are given.","level":"intermediate","tags":["benchmarks","evals","reasoning","agents","language-models"]}}

[Check it on Verify a thought](https://thesimind.si/proof-of-mind/)

## The authors

Francesco Belardinelli, Flavio Corradini, Matteo Leonesi

Credit for this research belongs to its authors. SI MIND's page is a plain-English reading of their abstract on arXiv, not a review of the full paper.

Related on the desk

## More papers on these topics

[Nov 18, 2026 · Number 7GPRM: Global Perspective Process Reward ModelGPRM is a step-by-step reasoning judge that sees earlier steps, their scores, and later steps, and the authors report it beats strong baselines on PRMBench and ProcessBench with less data and smaller models.ReasoningEvalsReinforcement learning](https://thesimind.si/papers/2611.20755/) · [Nov 17, 2026 · Number 10Adversarially Attacking Symbolic Vocabulary Vulnerabilities In LLM PlannersChanging the names of actions and objects in a planning task, using an automated attacker called Symbolic Swapper, can cut LLM planning performance by over 70%, even for strong reasoning models.AgentsSecurityReasoning](https://thesimind.si/papers/2611.17910/) · [Nov 19, 2026 · Number 12SciWorkBench: A Benchmark for Real Scientific WorkflowsSciWorkBench tests SI agents on real science tasks in four fields. The best agent scores 57.38, so current agents still struggle with realistic scientific workflows.BenchmarksAgentsScience](https://thesimind.si/papers/2611.22490/)

[Number 8](https://thesimind.si/papers/2611.26358/) · [All 12 for Nov 21, 2026](https://thesimind.si/papers/day/2026-11-21/) · [Number 10](https://thesimind.si/papers/2611.26601/)
