# Can Neural Networks Understand Logical Entailment?

arXiv: https://arxiv.org/abs/1802.08535 · HTML: https://mlsanity.com/paper/1802.08535/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Feb 2018.

## Built on by (most cited first, 24 of 24)

- [Proof Artifact Co-training for Theorem Proving with Language Models](https://mlsanity.com/paper/2102.06203/) (2021)
- [Machine Learning for Electronic Design Automation: A Survey](https://mlsanity.com/paper/2102.03357/) (2021)
- [LIME: Learning Inductive Bias for Primitives of Mathematical Reasoning](https://mlsanity.com/paper/2101.06223/) (2021)
- [Towards LogiGLUE: A Brief Survey and A Benchmark for Analyzing Logical Reasoning Capabilities of Language Models](https://mlsanity.com/paper/2310.00836/) (2023)
- [Logically Consistent Language Models via Neuro-Symbolic Integration](https://mlsanity.com/paper/2409.13724/) (2024)
- [Pushing the Limits of Rule Reasoning in Transformers through Natural Language Satisfiability](https://mlsanity.com/paper/2112.09054/) (2021)
- [Symbol-LLM: Leverage Language Models for Symbolic System in Visual Human Activity Reasoning](https://mlsanity.com/paper/2311.17365/) (2023)
- [Recognizing and Verifying Mathematical Equations using Multiplicative Differential Neural Units](https://mlsanity.com/paper/2104.02899/) (2021)
- [Generalized Quantifiers as a Source of Error in Multilingual NLU Benchmarks](https://mlsanity.com/paper/2204.10615/) (2022)
- [Learning Guided Automated Reasoning: A Brief Survey](https://mlsanity.com/paper/2403.04017/) (2024)
- [MizAR 60 for Mizar 50](https://mlsanity.com/paper/2303.06686/) (2023)
- [Generating Symbolic Reasoning Problems with Transformer GANs](https://mlsanity.com/paper/2110.10054/) (2021)
- [NL2LOGIC: AST-Guided Translation of Natural Language into First-Order Logic with Large Language Models](https://mlsanity.com/paper/2602.13237/) (2026)
- [Neuro-Symbolic VQA: A review from the perspective of AGI desiderata](https://mlsanity.com/paper/2104.06365/) (2021)
- [On the Capabilities of Pointer Networks for Deep Deductive Reasoning](https://mlsanity.com/paper/2106.09225/) (2021)
- [Giving AI Personalities Leads to More Human-Like Reasoning](https://mlsanity.com/paper/2502.14155/) (2025)
- [BAIT: Benchmarking (Embedding) Architectures for Interactive Theorem-Proving](https://mlsanity.com/paper/2403.03401/) (2024)
- [Embeddings as Epistemic States: Limitations on the Use of Pooling Operators for Accumulating Knowledge](https://mlsanity.com/paper/2210.05723/) (2022)
- [Zero-Shot Instruction Following in RL via Structured LTL Representations](https://mlsanity.com/paper/2602.14344/) (2026)
- [Learning Heuristics for Template-based CEGIS of Loop Invariants with Reinforcement Learning](https://mlsanity.com/paper/2107.09766/) (2021)
- [Extending Logical Neural Networks using First-Order Theories](https://mlsanity.com/paper/2207.02978/) (2022)
- [Structured Relational Representations](https://mlsanity.com/paper/2505.12143/) (2025)
- [Natural Language Satisfiability: Exploring the Problem Distribution and Evaluating Transformer-based Language Models](https://mlsanity.com/paper/2508.17153/) (2025)
- [Zero-Shot Instruction Following in RL via Structured LTL Representations](https://mlsanity.com/paper/2512.02633/) (2025)

## Abstract (arXiv)

We introduce a new dataset of logical entailments for the purpose of measuring models' ability to capture and exploit the structure of logical expressions against an entailment prediction task. We use this task to compare a series of architectures which are ubiquitous in the sequence-processing literature, in addition to a new model class---PossibleWorldNets---which computes entailment as a "convolution over possible worlds". Results show that convolutional networks present the wrong inductive bias for this class of problems relative to LSTM RNNs, tree-structured neural networks outperform LSTM RNNs due to their enhanced ability to exploit the syntax of logic, and PossibleWorldNets outperform all benchmarks.