Corpus ID: 236171337

Abstract Reasoning via Logic-guided Generation

  title={Abstract Reasoning via Logic-guided Generation},
  author={Sihyun Yu and Sangwoo Mo and Sungsoo Ahn and Jinwoo Shin},
  • Sihyun Yu, Sangwoo Mo, +1 author Jinwoo Shin
  • Published 2021
  • Computer Science
  • ArXiv
Reasoning via Logic-guided Generation Sihyun Yu 1 Sangwoo Mo 1 Sungsoo Ahn 2 Jinwoo Shin 1 

Figures and Tables from this paper


SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver
This paper introduces a differentiable (smoothed) maximum satisfiability (MAXSAT) solver that can be integrated into the loop of larger deep learning systems and demonstrates that by integrating this solver into end-to-end learning systems, this approach shows promise in integrating logical structures within deep learning. Expand
Abstract Reasoning with Distracting Features
This paper proposes feature robust abstract reasoning (FRAR) model, which consists of a reinforcement learning based teacher network to determine the sequence of training and a student network for predictions that is able to beat the state-of-the-art models. Expand
Abstract Diagrammatic Reasoning with Multiplex Graph Networks
This work proposes MXGNet, a multilayer graph neural network for multi-panel diagrammatic reasoning tasks that combines three powerful concepts, namely, object-level representation, graph neural networks and multiplex graphs, for solving visual reasoning tasks. Expand
The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning
This work proposes the Scattering Compositional Learner (SCL), an architecture that composes neural networks in a sequence that achieves state-of-the-art performance on two RPM datasets, and finds that the compositional representation makes the SCL significantly more robust to test-time domain shifts and greatly improves zero-shot generalization to previously unseen analogies. Expand
Stratified Rule-Aware Network for Abstract Visual Reasoning
A Stratified Rule-Aware Network (SRAN) is proposed to generate the rule embeddings for two input sequences and a rule similarity metric is applied to guarantee that SRAN can not only be trained using a tuplet loss but also infer the best answer efficiently. Expand
A simple neural network module for relational reasoning
This work shows how a deep learning architecture equipped with an RN module can implicitly discover and learn to reason about entities and their relations. Expand
Measuring abstract reasoning in neural networks
A dataset and challenge designed to probe abstract reasoning, inspired by a well-known human IQ test, is proposed and ways to both measure and induce stronger abstract reasoning in neural networks are introduced. Expand
RAVEN: A Dataset for Relational and Analogical Visual REasoNing
This work proposes a new dataset, built in the context of Raven's Progressive Matrices (RPM) and aimed at lifting machine intelligence by associating vision with structural, relational, and analogical reasoning in a hierarchical representation and establishes a semantic link between vision and reasoning by providing structure representation. Expand
Learning Perceptual Inference by Contrasting
It is demonstrated that CoPINet sets the new state-of-the-art for permutation-invariant models on two major datasets and concludes that spatial-temporal reasoning depends on envisaging the possibilities consistent with the relations between objects and can be solved from pixel-level inputs. Expand
Solve the problem first: constructive solution strategies can influence the accuracy of retrospective confidence judgments.
Two experiments tested whether differences in problem-solving strategies influence the ability of people to monitor their problem-Solving effectiveness as measured by confidence judgments and found that spontaneous constructive matching in nonverbal spatial reasoning problems was associated with better confidence calibration and resolution than response elimination. Expand