Opportunities for Co-Op and Industry-PhD Projects

Axiomatic_AI's mission:

Axiomatic_AI is launching with the aim to accelerate R&D by "Automated Interpretable Reasoning" (AIR) - a verifiably truthful AI model built for reasoning in science and engineering. Axiomatic_AI is hiring top talent interested in a future of human reasoning aided by - not replaced by - AI, and a future that empowers a new generation of innovators to solve important problems through deep-tech engineering in the semiconductor ecosystem.

Please see below for co-working project opportunities at Axiomatic_AI.

Competitive Programming Projects

Project 1: Enhancing AI-Powered Code Synthesis

Overview: This project focuses on advancing the capabilities of AI-powered code synthesis tools like AlphaCode. The goal is to develop algorithms that can automatically generate efficient and correct code from high-level problem descriptions.

• Objectives:
• Develop new algorithms for code generation that improve upon current state-of-the-art models.
• Implement a robust verification system to ensure the correctness of the generated code.
• Integrate the system with Axiomatic_AI's CDT generator and verifier.
• Expected Outcomes:
  • Enhanced code synthesis capabilities.
  • Improved accuracy and efficiency in generated code.
  • Seamless integration with Axiomatic_AI's existing platforms.
• Requirements: Strong background in machine learning, natural language processing, and programming languages.
• References:
• "AlphaCode: Developing Code Generation Algorithms" Research Paper
• GitHub - AlphaCode Repository, Codex Examples

Project 2: Optimizing Code through Automated Refactoring

Overview: This project aims to develop AI-driven tools for automated code refactoring, improving code quality and maintainability. The focus is on integrating these tools with Axiomatic_AI's suite of optimizers.

• Objectives:
• Create algorithms for identifying refactoring opportunities in codebases.
• Develop methods for automated code refactoring and optimization.
• Test and validate the tools within real-world code repositories.
• Expected Outcomes:
  • Automated tools for code refactoring.
  • Improved code quality and performance.
  • Integration with Axiomatic_AI's optimization suite.
• Requirements: Experience in software engineering, machine learning, and software optimization techniques.
• References:
• AlphaProof
• SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models : Ziyi Wu, Nikita Dvornik, Klaus Greff, Thomas Kipf, and Animesh Garg arXiv preprint arXiv:2210.05861 2022

Project : AI Code Synthesis for Microelectronics Design

Overview: This project focuses on developing AI-driven code synthesis tools for automating the design and verification of microelectronic circuits.

• Objectives:
• Develop AI algorithms for generating Verilog/VHDL code for microelectronics designs.
• Implement a verification system to ensure the correctness of synthesized designs.
• Test and validate the system on real-world microelectronics projects.
• Expected Outcomes:
  • Automated code synthesis tools for microelectronics design.
  • Improved design efficiency and correctness.
  • Validation on real-world microelectronics projects.
• Requirements: Strong background in digital circuit design, Verilog/VHDL, and machine learning.
• References:
• "Micro/Nano Circuits and Systems Design and Design Automation" Research Paper
• GitHub - OpenROAD: Open Source EDA
• https://arxiv.org/abs/2405.16380
• GitHub - EDA Tools and Resources

Project : AI-Driven Photonic Integrated Circuit (PIC) Design Automation

Overview: This project aims to create AI-powered tools for the design and optimization of photonic integrated circuits (PICs), enhancing the design process and reducing time-to-market.

• Objectives:
• Develop AI algorithms for synthesizing PIC designs from high-level specifications.
• Create optimization techniques for improving PIC performance and efficiency.
• Validate the tools with real-world PIC designs.
• Expected Outcomes:
  • AI-driven synthesis tools for PIC design.
  • Enhanced performance and efficiency of PICs.
  • Successful validation with real-world PIC projects.
• Requirements: Expertise in photonic circuit design, optimization algorithms, and machine learning.
• References:
• https://proceedings.mlr.press/v235/chen24ad.html
• GitHub - Photonics Simulation Tools

Digital Twins Projects

Project 1: Advanced Digital Twin Integration for AXI

Overview: This project explores the integration of digital twin technologies within the Axiomatic_AI framework, focusing on real-time data synchronization and predictive analytics.

• Objectives:
• Develop methods for real-time data integration from IoT devices into digital twins.
• Implement predictive analytics to enhance operational efficiency.
• Validate the system in AXI-relevant industries.
• Expected Outcomes:
  • Real-time integrated digital twin systems.
  • Enhanced predictive analytics capabilities.
  • Demonstrated benefits in AXI-relevant industries.
• Requirements: Background in IoT, data analytics, and digital twin technologies.
• References:
• "Digital Twin for Industry 4.0: Real-Time Integration and Analytics" Research Paper
• GitHub - Azure Digital Twins
• "Predictive Analytics in Industry 4.0 Using Digital Twins" Research Paper
• GitHub - Industry 4.0 Solutions

Project 2: Digital Twin Framework for Engineering Systems

Overview: This project focuses on creating a comprehensive digital twin framework for engineering systems, enabling better design, simulation, and validation processes.

• Objectives:
• Develop a scalable framework for creating digital twins of engineering systems.
• Integrate real-time data from various engineering processes.
• Implement analytics for design and operational optimization.
• Expected Outcomes:
  • Scalable digital twin framework for engineering systems.
  • Enhanced design and operational optimization capabilities.
  • Successful pilot deployment in AXI-relevant engineering projects.
• Requirements: Expertise in engineering design, data integration, and digital twin technologies.
• References:
• https://nap.nationalacademies.org/catalog/26894/foundational-research-gaps-and-future-directions-for-digital-twins?utm_source=NASEM+Math+and+Statistics&utm_campaign=87b2f564c2-EMAIL_CAMPAIGN_2023_05_15_01_42_COPY_01&utm_medium=email&utm_term=0_-a0739a5cef-%5BLIST_EMAIL_ID%5D
• NVIDIA Omniverse

Probabilistic Machine Learning Projects

Project 1: Probabilistic Models for Uncertainty Quantification in AI

Overview: This project aims to develop probabilistic models that can quantify uncertainty in AI predictions, improving the reliability of AI systems.

• Objectives:
• Develop new probabilistic models for uncertainty quantification.
• Integrate these models with existing AI systems to enhance decision-making.
• Validate the models in real-world applications.
• Expected Outcomes:
  • Improved uncertainty quantification models.
  • Enhanced reliability of AI predictions.
  • Successful integration and validation in real-world scenarios.
• Requirements: Strong background in probabilistic modeling, statistics, and machine learning.
• References:
• https://probml.github.io/pml-book/book1.html
• GitHub - Bayesian Deep Learning

Project 2: Integrating Factor Networks and Knowledge Graphs for Enhanced AI Reasoning

Overview: This project explores the use of factor networks and knowledge graphs to improve AI reasoning and decision-making processes.

• Objectives:
• Develop methods for integrating factor networks with knowledge graphs to represent complex relationships.
• Apply these integrated models to enhance AI reasoning and inference capabilities.
• Validate the effectiveness of the integrated models in real-world scenarios.
• Expected Outcomes:
  • Advanced techniques for integrating factor networks and knowledge graphs.
  • Improved AI reasoning and decision-making capabilities.
  • Validation through case studies in various domains.
• Requirements: Expertise in probabilistic graphical models, knowledge graphs, and machine learning.
• References:
• "Knowledge Graphs: Principles and Applications" Research Paper
• GitHub - Knowledge Graph Toolkit