Towards Improving Reasoning & Planning Capabilities of LLMs with Neuro-Symbolic Learning

✨✨ Curated collection of papers and resources on latest advances on improving reasoning and planning abilities of LLM/MLLMs with neuro-symbolic learning

🗂️ Table of Contents

• Awesome Tutorials & Workshops & Talks
• Awesome Survey
• Basic Neuro-Symbolic Frameworks
• Symbolic to LLM
• Symbolic Generation, LLM Imitation
• LLM Formalize, Symbolic Augment
• LLM to Symbolic
• Symbolic Solver Aided Methods
• Program Aided Methods
• Tool Aided Methods
• Search Augmented Methods
• LLM plus Symbolic
• Symbolic Formatted Reasoning
• Differential Symbolic Module
• Symbolic Feedback
• Awesome Datasets & Benchmarks

📌 Awesome Tutorials & Workshops & Talks

• Neuro-Symbolic Visual Reasoning and Program Synthesis Tutorials in CVPR 2020
• Neuro-Symbolic Methods for Language and Vision Tutorials in AAAI 2022
• AAAI 2022 Tutorial on AI Planning: Theory and Practice Tutorials in AAAI 2022
• Advances in Neuro Symbolic Reasoning and Learning Tutorials in AAAI 2023
• Neuro-Symbolic Approaches: Large Language Models + Tool Use Tutorials in ACL 2023
• Neuro-Symbolic Generative Models Workshop in ICLR 2023
• Neuro-Symbolic Learning and Reasoning in the Era of Large Language Models Workshop in AAAI 2024
• Neuro-Symbolic Concepts for Robotic Manipulation Talk given by Jiayuan Mao [[Video]](https://www.youtube.com/watch?v=S8KsCtbJqz0)
• Building General-Purpose Robots with Compositional Action Abstractions Talk given by Jiayuan Mao
• Summer School on Neurosymbolic Programming
• MIT 6.S191: Neuro-Symbolic AI Talk given by David Cox [[Video]](https://www.youtube.com/watch?v=4PuuziOgSU4)
• NeuroSymbolic Programming [[Slides]](https://nips.cc/media/neurips-2022/Slides/55804.pdf)
• LLM Reasoning: Key Ideas and Limitations Talk give by Denny Zhou
• Inference-Time Techniques for LLM Reasoning Talk given by Xinyun Chen
• Neurosymbolic Reasoning for Large Language Models Neuro-Symbolic AI Summer School in UCLA, 2024

🔍 Survey

Survey on LLM Reasoning

• Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models
• LLM Post-Training: A Deep Dive into Reasoning Large Language Models
• A Survey on Post-training of Large Language Models
• Reasoning Language Models: A Blueprint
• Towards Large Reasoning Models: A Survey of Reinforced Reasoning with Large Language Models
• Logical Reasoning in Large Language Models: A Survey
• From System 1 to System 2: A Survey of Reasoning Large Language Models
• A Survey on LLM Inference-Time Self-Improvement
• Empowering LLMs with Logical Reasoning: A Comprehensive Survey
• Advancing Reasoning in Large Language Models: Promising Methods and Approaches
• A Survey on Deep Learning for Theorem Proving
• A Survey of Mathematical Reasoning in the Era of Multi-Modal Large Language Model: Benchmark, Method & Challenges
• Multi-Modal Chain-of-Thought Reasoning:A Comprehensive Survey
• Exploring the Reasoning Abilities of Multi-Modal Large Language Models (MLLMs): A Comprehensive Survey on Emerging Trends in Multimodal Reasoning

Survey on LLM Planning

• A Survey on Large Language Models for Automated Planning
• A Survey of Optimization-based Task and Motion Planning: From Classical To Learning Approaches
• A Survey on Large Language Model based Autonomous Agents
• Understanding the planning of LLM agents: A survey
• Introduction to AI Planning

Survey on Neuro-Symbolic Learning

• A Survey on Neural-symbolic Learning Systems
• Towards Cognitive AI Systems: a Survey and Prospective on Neuro-Symbolic AI
• Bridging the Gap: Representation Spaces in Neuro-Symbolic AI
• Neuro-Symbolic AI: The 3rd Wave
• Neuro-Symbolic AI and its Taxonomy: A Survey
• The third AI summer: AAAI Robert S. Engelmore Memorial Lecture
• NeuroSymbolic AI - Why, What, and How
• From Statistical Relational to Neuro-Symbolic Artificial Intelligence: a Survey
• Neuro-Symbolic Artificial Intelligence: Current Trends
• Neuro-Symbolic Reinforcement Learning and Planning: A Survey
• A Review on Neuro-symbolic AI Improvements to Natural Language Processing
• Survey on Applications of NeuroSymbolic Artificial Intelligence
• Overview of Neuro-Symbolic Integration Frameworks

📖 Basic Neuro-Symbolic Frameworks

| Title | Venue | Date | Code | |:--------|:--------:|:--------:|:--------:| |Semantic-based regularization for learning and inference
| Artificial Intelligence | 2017 | - | |DeepProbLog: Neural Probabilistic Logic Programming
| NeurIPS | 2018 | Github | |Learning Explanatory Rules from Noisy Data
| Journal of Artificial Intelligence Research | 2018 | Github | |Augmenting Neural Networks with First-order Logic
| ACL | 2019 | - | |Neural Logic Machines
| ICLR | 2019 | Github | |Bridging Machine Learning and Logical Reasoning by Abductive Learning
| NeurIPS | 2019 | Github | |SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver
| ICML | 2019 | Github | |DL2: Training and Querying Neural Networks with Logic
| ICML | 2019 | Github | |The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision
| ICLR | 2019 | Github | |NeurASP: Embracing Neural Networks into Answer Set Programming
| IJCAI | 2020 | Github | |Learning programs by learning from failures
| Machine Learning | 2020 | - | |Logical Neural Networks
| Arxiv | 2020 | Github | |Closed Loop Neural-Symbolic Learning via Integrating Neural Perception, Grammar Parsing, and Symbolic Reasoning
| ICML | 2020 | Github | |Ontology Reasoning with Deep Neural Networks
| Artificial Intelligence | 2020 | - | |MultiplexNet: Towards Fully Satisfied Logical Constraints in Neural Networks
| AAAI | 2022 | - | |Neuro-Symbolic Hierarchical Rule Induction
| ICML | 2022 | - | |Logic Tensor Networks
| Artificial Intelligence | 2022 | Github | |Neuro-symbolic Learning Yielding Logical Constraints
| NeurIPS | 2023 | Github | |Neural-Symbolic Recursive Machine for Systematic Generalization
| ICLR | 2024 | Github |

📖 Symbolic to LLM

Symbolic Generation, LLM Imitation

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |Solving Olympiad geometry without human demonstrations| Nature | 2024 |Geometric Math Reasoning|Github| |Chain of Thought Imitation with Procedure Cloning
| NeurIPS | 2022 |Planning|Github| |Plansformer: Generating Symbolic Plans using Transformers
| Arxiv | 2022 |Planning|-| |Beyond A$^*$: Better Planning with Transformers via Search Dynamics Bootstrapping| Arxiv | 2024 |Planning|Github| |Dualformer: Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Traces
| Arxiv | 2024 |Planning|Github| |Stream of Search (SoS): Learning to Search in Language
| COLM | 2024 |Reasoning|Github| |Step Back to Leap Forward: Self-Backtracking for Boosting Reasoning of Language Models|Arxiv| 2025 |Reasoning|Github| |Language Models can be Deductive Solvers
| NAACL| 2024 |Logical Reasoning| Github | |INT: An Inequality Benchmark for Evaluating Generalization in Theorem Proving|ICLR| 2021 | Theorem Proving |Github| |Generating Millions Of Lean Theorems With Proofs By Exploring State Transition Graphs|Arxiv| 2025 | Theorem Proving |-| |Enhancing Reasoning Capabilities of LLMs via Principled Synthetic Logic Corpus
| NeurIPS | 2024 |Logical Reasoning|-| |A Symbolic Framework for Evaluating Mathematical Reasoning and Generalization with Transformers
| NAACL | 2024 |Math Reasoning|Github| |Proving Olympiad Algebraic Inequalities without Human Demonstrations
| NeurIPS | 2024 |Theorem Proving| Github | |LogicAsker: Evaluating and Improving the Logical Reasoning Ability of Large Language Models
| EMNLP | 2024 |Logical Reasoning| Github |

LLM Formalize, Symbolic Augment

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |AMR-DA: Data Augmentation by Abstract Meaning Representation
| ACL| 2022 |Logic Reasoning|Github | |Abstract Meaning Representation-Based Logic-Driven Data Augmentation for Logical Reasoning
| ACL| 2024 |Logic Reasoning|Github | |Neuro-Symbolic Data Generation for Math Reasoning
| NeurIPS | 2024 |Math Reasoning|-| |LawGPT: Knowledge-Guided Data Generation and Its Application to Legal LLM
| SCI-FM Workshop @ ICLR | 2025 |Legal Reasoning|Github| |AlphaIntegrator: Transformer Action Search for Symbolic Integration Proofs
| Arxiv| 2024 |Theorem Proving| - | |Large Language Models Meet Symbolic Provers for Logical Reasoning Evaluation
| ICLR | 2025 |Logic Reasoning| Github|

📖 LLM to Symbolic

Symbolic Solver Aided Methods

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |MRKL Systems: A modular, Neuro-Symbolic Architecture that Combines Large Language models, External Knowledge Sources and Discrete Reasoning
| Arxiv | 2022 |Reasoning| - | |Leveraging Large Language Models to Generate Answer Set Programs
| KR | 2023 |Reasoning|Github| |Logic-LM: Empowering Large Language Models with Symbolic Solvers for Faithful Logical Reasoning
| EMNLP | 2023 |Reasoning|Github| |LOGIC-LM++: Multi-Step Refinement for Symbolic Formulations
| ACL | 2024 |Reasoning|-| |LINC: A Neurosymbolic Approach for Logical Reasoning by Combining Language Models with First-Order Logic Provers
| EMNLP | 2023 |Reasoning|Github| |Faithful Chain-of-Thought Reasoning
| IJCNLP-AACL | 2023 |Reasoning|Github| |SATLM: Satisfiability-Aided Language Models Using Declarative Prompting
| NeurIPS | 2023 |Math Reasoning|Github| |Faithful Logical Reasoning via Symbolic Chain-of-Thought
| ACL | 2023 |Reasoning|Github| |Verification and Refinement of Natural Language Explanations through LLM-Symbolic Theorem Proving
| EMNLP | 2024 |Theorem Proving| Github | |StackSight: Unveiling WebAssembly through Large Language Models and Neurosymbolic Chain-of-Thought Decompilation
| ICML | 2024 |Code Generation|-| |Monitor-Guided Decoding of Code LMs with Static Analysis of Repository Context
| NeurIPS | 2023 |Code Generation|Github| |Solving Math Word Problems by Combining Language Models With Symbolic Solvers
| Arxiv | 2023 |Math Reasoning|Github| |AutoSAT: Automatically Optimize SAT Solvers via Large Language Models
| Arxiv | 2024 |SAT Problem|Github| |Prototype-then-Refine: A Neuro-Symbolic Approach for Improved Logical Reasoning with LLMs
| Arxiv | 2024 |Reasoning|-| |Neuro-Symbolic Integration Brings Causal and Reliable Reasoning Proofs
| Arxiv | 2024 |Reasoning| Github | |SymBa: Symbolic Backward Chaining for Structured Natural Language Reasoning
| Arxiv | 2025 |Reasoning|Github| |Frugal LMs Trained to Invoke Symbolic Solvers Achieve Parameter-Efficient Arithmetic Reasoning
| AAAI | 2024 |Math Reasoning|-| |Lemur: Integrating Large Language Models in Automated Program Verification
| ICLR | 2024 |Code Generation|-| |Symbol-LLM: Leverage Language Models for Symbolic System in Visual Human Activity Reasoning
| NeurIPS | 2023 |Robotics|Github| |Parsel: Algorithmic Reasoning with Language Models by Composing Decompositions
| NeurIPS | 2023 |Robotics| Github | |Disentangling Extraction and Reasoning in Multi-hop Spatial Reasoning
| EMNLP | 2024 |Spatial Reasoning|-| |Generalized Planning in PDDL Domains with Pretrained Large Language Models
| AAAI | 2024 |Planning|Github| |Leveraging Pre-trained Large Language Models to Construct and Utilize World Models for Model-based Task Planning
| NeurIPS | 2023 |Planning|Github| |Coupling Large Language Models with Logic Programming for Robust and General Reasoning from Text
| ACL Findings | 2023 |Planning|Github| |LLM+P: Empowering Large Language Models with Optimal Planning Proficiency
| Arxiv | 2023 |Planning/Robotics|Github| |Dynamic Planning with a LLM
| LanGame Workshop @ NeurIPS | 2023 |Planning/Robotics|-| |Neuro-Symbolic Procedural Planning with Commonsense Prompting
| ICLR | 2023 |Planning/Robotics|Github| |Leveraging Environment Interaction for Automated PDDL Translation and Planning with Large Language Models
| NeurIPS | 2024 | Planning |Github | |A Framework for Neurosymbolic Robot Action Planning using Large Language Models
| Frontiers in Neurorobotics | 2024 |Robotics|Github| |ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation
| Arxiv | 2024 |Robotics|-| |LeanDojo: Theorem Proving with Retrieval-Augmented Language Models
| NeurIPS D&B | 2023 |Theorem Proving| Github | |LEGO-Prover: Neural Theorem Proving with Growing Libraries
| ICLR | 2024 |Theorem Proving| Github | |Proving Olympiad Inequalities by Synergizing LLMs and Symbolic Reasoning
| ICLR | 2025 |Theorem Proving|Github| |Autoformalization with Large Language Models
| NeurIPS | 2022 |Theorem Proving| - | |Autoformalizing Euclidean Geometry
| ICML | 2024 | Geometry Reasoning|Github | |Autoformalize Mathematical Statements by Symbolic Equivalence and Semantic Consistency
| NeurIPS | 2024 |Theorem Proving|Github| |Draft, Sketch, and Prove: Guiding Formal Theorem Provers with Informal Proofs
| ICLR | 2023 |Theorem Proving|Github| |Don't Trust: Verify -- Grounding LLM Quantitative Reasoning with Autoformalization
| ICLR | 2024 |Theorem Proving|Github| |Large Language Models as Planning Domain Generators
| ICAPS | 2024 |Planning|Github| |Generating Symbolic World Models via Test-time Scaling of Large Language Models
| ICML | 2024 |Planning|-|

Program-Aided Methods

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |PAL: Program-aided Language Models
| ICML | 2023 | Reasoning |Github | |Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks
| TMLR | 2023 |Math Reasoning| Github | |Binding Language Models in Symbolic Languages
| ICLR | 2023 |Reasoning|Github| |Chain of Code: Reasoning with a Language Model-Augmented Code Emulator|ICML| 2024 |Reasoning|Github| |CODE4STRUCT: Code Generation for Few-Shot Event Structure Prediction|ACL| 2023 |Reasoning|Github| |MathPrompter: Mathematical Reasoning using Large Language Models|ACL| 2023 |Math Reasoning|Github| |Natural Language Embedded Programs for Hybrid Language Symbolic Reasoning
| ACL | 2024 |Reasoning| Github | |Towards Better Understanding of Program-of-Thought Reasoning in Cross-Lingual and Multilingual Environments|Arxiv| 2025 |Reasoning|-| |Code as Policies: Language Model Programs for Embodied Control
| Arxiv | 2023 |Robotics|Github|

Tool-Aided Methods

| Title | Venue | Date | Code | |:--------|:--------:|:--------:|:--------:| |Visual Programming: Compositional visual reasoning without training
| CVPR (Best Paper) | 2023 |Visual Reasoning| Github | |ViperGPT: Visual Inference via Python Execution for Reasoning
| ICCV | 2023 |Visual Reasoning|Github | |GENOME: Generative Neuro-Symbolic Visual Reasoning by Growing and Reusing Modules
| ICLR | 2024 |Visual Reasoning|-| |Chameleon: Plug-and-Play Compositional Reasoning with Large Language Models
| NeurIPS | 2023 |Multi-Modal Reasoning|Github| |HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face
| NeurIPS| 2023 |Visual Reasoning|-| |ART: Automatic Multi-Step Reasoning and Tool-use for Large Language Models
| Arxiv | 2023 |Reasoning|-| |ChatCoT: Tool-Augmented Chain-of-Thought Reasoning on Chat-based Large Language Models
| EMNLP | 2023 |Reasoning| Github | |ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solving
| ICLR | 2024 |Math Reasoning|Github| |ToolkenGPT: Augmenting Frozen Language Models with Massive Tools via Tool Embeddings
| NeurIPS | 2023 |Agent|Github| |TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs
| Arxiv| 2023 |Agent|-| |ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
|ICLR| 2024 |Agent|Github| |Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool Usage
|ICLR| 2025 |Visual Reasoning|Github| |ToolPlanner: A Tool Augmented LLM for Multi Granularity Instructions with Path Planning and Feedback
|EMNLP| 2024 |Reasoning|Github| |START: Self-taught Reasoner with Tools
|Arxiv| 2025 |Reasoning|-|

Search Augmented Methods

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |Self-Evaluation Guided Beam Search for Reasoning
| NeurIPS | 2023 |Reasoning| Github | |Deductive Beam Search: Decoding Deducible Rationale for Chain-of-Thought Reasoning
| COLM | 2024 |Reasoning| Github | |MindStar: Enhancing Math Reasoning in Pre-trained LLMs at Inference Time
| Arxiv | 2024 |Reasoning|-| |NEUROLOGIC A*esque Decoding:Constrained Text Generation with Lookahead Heuristics
| NAACL | 2022 |Reasoning|-| |Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions
| Arxiv | 2024 |Reasoning|Github| |Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning
| Arxiv | 2024 |Reasoning| Github | |LLaMA-Berry: Pairwise Optimization for Olympiad-level Mathematical Reasoning via O1-like Monte Carlo Tree Search
| Arxiv | 2024 |Reasoning| Github | |Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Search
| Arxiv | 2024 |Visual Reasoning|Github| |Large Language Models as Commonsense Knowledge for Large-Scale Task Planning
| NeurIPS | 2023 |Planning| Github | |Planning with Large Language Models for Code Generation
| ICLR | 2023 |Code Generation| Github | |Tree of Thoughts: Deliberate Problem Solving with Large Language Models
| NeurIPS | 2023 |Reasoning| Github | |CodeTree: Agent-guided Tree Search for Code Generation with Large Language Models
| Arxiv | 2024 |Code Generation|-| |Alphazero-like Tree-Search can Guide Large Language Model Decoding and Training
| Arxiv | 2024 |Reasoning|-| |Reasoning with Language Model is Planning with World Model
| EMNLP | 2023 |Reasoning|-| |Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning
| ICLR | 2024 |Reasoning| - | |SPaR: Self-Play with Tree-Search Refinement to Improve Instruction-Following in Large Language Models
| ICLR | 2025 |Reasoning| Github | |Leveraging Constrained Monte Carlo Tree Search to Generate Reliable Long Chain-of-Thought for Mathematical Reasoning
| Arxiv | 2025 |Math Reasoning|-| |FGeo-DRL:Deductive Reasoning for Geometric Problems through Deep Reinforcement Learning
| Arxiv | 2024 |Geometry Reasoning|Github| |Interactive Evolution: A Neural-Symbolic Self-Training Framework For Large Language Models|Arxiv|2024|Reasoning|Github| |AlphaMath Almost Zero: Process Supervision without Process|NeurIPS| 2024 |Math Reasoning|Github| |Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotations
| ACL | 2024 |Math Reasoning|Github| |Improve Mathematical Reasoning in Language Models by Automated Process Supervision| DeepMind Report| 2024 |Math Reasoning|-| |ReST-MCTS: LLM Self-Training via Process Reward Guided Tree Search| NeurIPS | 2024 |Math/Chemistry/Physics|Github| |rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking
| ICLR | 2025 |Math Reasoning| Github | |Enhancing Reasoning through Process Supervision with Monte Carlo Tree Search| NeurMAD Workshop @ AAAI| 2025 |Math Reasoning|-| |STP: Self-play LLM Theorem Provers with Iterative Conjecturing and Proving|Arxiv| 2025 |Theorem Proving|Github| |SRA-MCTS: Self-driven Reasoning Augmentation with Monte Carlo Tree Search for Code Generation|Arxiv| 2024 |Reasoning|Github|

📖 LLM plus Symbolic

Symbolic Formatted Reasoning

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |Deductive Verification of Chain-of-Thought Reasoning|NeurIPS| 2023 |Reasoning|Github| |Show Your Work: Scratchpads for Intermediate Computation with Language Models
| DL4C Workshop @ ICLR | 2022 |Code Generation|-| |Chain-of-Symbol Prompting Elicits Planning in Large Langauge Models|COLM| 2024 |Spatial Reasoning/Path Planning|Github| |Learning to Reason via Program Generation, Emulation, and Search|NeurIPS| 2024 |Reasoning|Github| |Symbolic Working Memory Enhances Language Models for Complex Rule Application|EMNLP| 2024 |Reasoning|Github| |CodeI/O: Condensing Reasoning Patterns via Code Input-Output Prediction|Arxiv| 2025 |Reasoning|Github| |SKIntern: Internalizing Symbolic Knowledge for Distilling Better CoT Capabilities into Small Language Models|COLING| 2025 |Reasoning|Github| |Improving Chain-of-Thought Reasoning via Quasi-Symbolic Abstractions|Arxiv| 2025 |Reasoning|-| |ProgPrompt: Generating Situated Robot Task Plans using Large Language Models
| ICRA | 2023 |Robotics|Github| |Programmatically Grounded, Compositionally Generalizable Robotic Manipulation
| ICLR | 2023 |Robotics|Github| |CodePlan: Unlocking Reasoning Potential in Large Language Models by Scaling Code-form Planning
| ICLR | 2025 |Reasoning|Github|

Differential Symbolic Modules

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |Inferring and Executing Programs for Visual Reasoning
| ICCV | 2017 |Visual Reasoning|Github| |Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding
| NeurIPS | 2018 |Visual Reasoning|Github| |The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision
| ICLR | 2019 |Visual Reasoning| Github | |Visual Concept-Metaconcept Learning
| NeurIPS | 2019 |Visual Reasoning| Github | |Neuro-Symbolic Visual Reasoning: Disentangling “Visual” from “Reasoning”
| ICML | 2020 |Visual Reasoning|Github| |Dynamic Visual Reasoning by Learning Differentiable Physics Models from Video and Language
| NeurIPS | 2021 |Visual Reasoning|Github| |Grounding Physical Concepts of Objects and Events Through Dynamic Visual Reasoning
| ICLR | 2021 |Visual Reasoning|Github| |JARVIS: A Neuro-Symbolic Commonsense Reasoning Framework for Conversational Embodied Agents
| Arxiv | 2022 |Robotics| - | |FALCON: Fast Visual Concept Learning by Integrating Images, Linguistic descriptions, and Conceptual Relations
| ICLR | 2022 |Visual Reasoning| Github| |NS3D: Neuro-Symbolic Grounding of 3D Objects and Relations
| CVPR | 2023 | Visual Reasoning |Github | |Interpretable Neural-Symbolic Concept Reasoning
| ICML | 2023 | Interpretable |Github | |Motion Question Answering via Modular Motion Programs
| ICML | 2023 | Motion QA |Github | |Learning Neuro-symbolic Programs for Language Guided Robot Manipulation
| ICRA | 2023 | Robotics |Github | |\alpha ILP: thinking visual scenes as differentiable logic programs
| Machine Learning | 2023 | Visual Reasoning | | |What's Left? Concept Grounding with Logic-Enhanced Foundation Models
| NeurIPS | 2023 | Visual Reasoning |Github | |Programmatically Grounded, Compositionally Generalizable Robotic Manipulation
| ICLR | 2023 |Robotics|Github| |Neuro-Symbolic Predicate Invention:Learning Relational Concepts from Visual Scenes
| NSAI | 2024 | Visual Reasoning |Github | |Take A Step Back: Rethinking the Two Stages in Visual Reasoning
| ECCV | 2024 | Visual Reasoning |Github | |Learning Differentiable Logic Programs for Abstract Visual Reasoning
| MLJ | 2024 |Visual Reasoning|-| |DiLA: Enhancing LLM Tool Learning with Differential Logic Layer
| Arxiv | 2024 |Reasoning|-| |Mastering Symbolic Operations: Augmenting Language Models with Compiled Neural Networks
| ICLR | 2024 |Reasoning| Github | |Empowering Language Models with Knowledge Graph Reasoning for Question Answering
| EMNLP | 2022 |Reasoning|-| |Neuro-symbolic Training for Spatial Reasoning over Natural Language
| Arxiv | 2025 |Spatial Reasoning|Github| |NeSyCoCo: A Neuro-Symbolic Concept Composer for Compositional Generalization
| Arxiv | 2024 |Visual Reasoning|Github|

Symbolic Feedback

| Title | Venue | Date | Domain | Code | |:--------|:--------:|:--------:|:--------:|:--------:| |CoTran: An LLM-based Code Translator using Reinforcement Learning with Feedback from Compiler and Symbolic Execution
| ECAI | 2024 |Code Generation|-| |Position: LLMs Can’t Plan, But Can Help Planning in LLM-Modulo Frameworks
| ICML | 2024 |Planning|-| |RLSF: Reinforcement Learning via Symbolic Feedback
| Arxiv | 2025 |Reasoning|Github| |Rule Based Rewards for Language Model Safety
| NeurIPS | 2024 | - |

Misc on Neuro-Symbolic Learning

| Title | Venue | Date | Code | |:--------|:--------:|:--------:|:--------:| |Neuro-Symbolic Entropy Regularization
| UAI | 2022 |Github| |RuleMatch: Matching Abstract Rules for Semi-supervised Learning of Human Standard Intelligence Tests
| IJCAI | 2023 |Github| |Learning with Logical Constraints but without Shortcut Satisfaction
| ICLR | 2023 |Github| |Neuro-Symbolic Continual Learning:Knowledge, Reasoning Shortcuts and Concept Rehearsal
| ICML | 2023 | Github | |Out-of-Distribution Generalization by Neural-Symbolic Joint Training
| AAAI | 2023 |Github| |Not All Neuro-Symbolic Concepts Are Created Equal: Analysis and Mitigation of Reasoning Shortcuts
| NeurIPS | 2023 |-| |Localized Symbolic Knowledge Distillation for Visual Commonsense Models
| NeurIPS | 2023 |-| |A-NeSI: A Scalable Approximate Method for Probabilistic Neurosymbolic Inference
| NeurIPS | 2023 |-| |Neuro-Symbolic Continual Learning:Knowledge, Reasoning Shortcuts and Concept Rehearsal
| ICML | 2023 | Github | |Out-of-Distribution Generalization by Neural-Symbolic Joint Training
| AAAI | 2023 |Github| |Large Language Models Are NeuroSymbolic Reasoners
| AAAI | 2024 |Github | |On the Hardness of Probabilistic Neurosymbolic Learning
| ICML | 2024 |-| |On the Independence Assumption in Neurosymbolic Learning
| ICML | 2024 |-| |Analysis for Abductive Learning and Neural-Symbolic Reasoning Shortcuts
| ICML | 2024 |-| |Convex and Bilevel Optimization for Neural-Symbolic Inference and Learning
| ICML | 2024 |-| |Bridging Neural and Symbolic Representations with Transitional Dictionary Learning
| ICLR | 2024 |-| |LogicMP: A Neuro-symbolic Approach for Encoding First-order Logic Constraints
| ICLR | 2024 |-| |Adaptable Logical Control for Large Language Models
| NeurIPS | 2024 | - | |Rule Extrapolation in Language Models: A Study of Compositional Generalization on OOD Prompts
| NeurIPS | 2024 | - | |Unlocking the Capabilities of Thought: A Reasoning Boundary Framework to Quantify and Optimize Chain-of-Thought
| NeurIPS | 2024 | Github|

🛠️ Awesome Datasets & Benchmarks

Mathematical Reasoning

• GSM8K, MATH, AIME, OlympiadBench, MiniF2F, GSM Symbolic, MWPBench, AMC, AddSub, MathQA, FIMO, TRIGO, U-MATH, Mario, MultiArith, CHAMP, ARB, LeanDojo, LISA, PISA, TheoremQA, FrontierMath, Functional, TABMWP, SCIBENCH, MultiHiertt, ChartQA

Logical Reasoning

• LogicGame, LogiQA, LogiQA-v2.0, PrOntoQA, ProofWriter, BigBench, FOLIO, AbductionRules, ARC Challenge, WANLI, CLUTRR, Adversarial NLI, Adversarial ARCT

Code Generation

• DS-1000, SWE-bench, APPS, MBPP, HumanEval, LiveCodeBench, ODEX

Visual Reasoning

• Visual Sudoku, CLEVR Dataset, GQA Dataset, VQA & VQA v2.0, Flickr30k entities, DAQUAR, Visual Genome, Visual7W, COCO-QA, TDIUC, SHAPES, VQA-Rephrasings, VQA P2, VQA-HAT, VQA-X, VQA-E, TallyQA, ST-VQA, Text-VQA, FVQA, OK-VQA

Geometry Reasoning

• Geometry3K, GEOQA, UniGEO, MathVista

Classical Planning

• PlanBench,LogiCity, Minigrid, BabyAI, Minecraft

Game AI Planning

• Atari 100k, Procgen, Gym Retro, Malmö, Obstacle Tower, Torcs, DeepMind Lab, Hard Eight, DeepMind Control, VizDoom, Pommerman, Multiagent emergence, Google Research Football, Neural MMOs, StarCraft II, PySC2, Fever Basketball

Robotic Planning

• Mini-Behavior, CLIPort Dataset, ALFworld, VirtualHome, RocoBench, Behavior, SMART-LLM, PPNL, Robotouille

AI Agent Planning

• WebArena, OSWorld, API-Bank, TravelPlanner, ChinaTravel, TaskBench, WebShop, AgentBench, AgentGym, AgentBoard, GAIA, MINT

Others

• RSbench: A Neuro-Symbolic Benchmark Suite for Concept Quality and Reasoning Shortcuts

↑ Back to Top ↑