Our goal is to develop an open-source AI model capable of complex mathematics and detailed data analysis, enhanced by incentivized human feedback for continuous improvement.
Welcome to LogicNet, where we’re building a specialized model for the computation of complex mathematics, logical reasoning (computational thinking) and data analysis.
Sophisticated language processing technologies, such as ChatGPT, Microsoft Copilot, and Google Gemini, never cease to impress with their range of functions. Despite their impressive abilities, they are not flawless. One significant issue they face is the generation of inaccurate information, a challenge that their developers are actively working to overcome. Nonetheless, my perspective is that the evolution of artificial intelligence will favour more focused, niche models over these broad-spectrum variants.
Links: Github | Discord Channel
Tony Le (AIT Protocol, CTO)
Co-founder of PharmApp Tech Inc.
Solutions designed with more than 60 data-driven strategies, incorporating machine learning to enhance profitability, efficiency, and business success
MIT alumni in Data Science & Machine Learning program
Mathematics for Machine Learning Specialization, Imperial College London
SQL for Data Science Specialization, University California, Davis
Google Cloud Professional Machine Learning Engineer
Google Cloud Professional Data Engineer
SnowPro Core Certification
Detailed Profile: See Tony’s LinkedIn for more info
Bittensor Guru Podcast: Listen to Tony’s interview about our subnet plans
LogicNet is a marketplace for AI/ML models specializing in mathematics, logical reasoning, and data analysis. It will act as a supercharger for other LLMs on Bittensor, by minimizing hallucinations in LLMs when asked to perform complex mathematical tasks. We achieve this by enabling the language model to autonomously write, test, and execute code within unique Python environments.
By providing a model capable of independent code writing and execution, we bolster the capabilities of miners and other subnets, thereby enhancing their accuracy and improving the quality of responses network-wide.
Our subnet contribution is pivotal, transcending direct user support to amplify the collective intelligence of the Bittensor ecosystem.
AI systems that can solve complex math could allow us to build more powerful AI tools.
Math is really, really hard for AI models. Complex math, such as geometry, requires sophisticated reasoning skills, and many AI researchers believe that the ability to crack it could herald more powerful and intelligent systems.
We believe LogicNet subnet can move Bittensor LLMs closer to human-like reasoning skills. This could allow us to build more powerful AI tools on Bittensor to help mathematicians solve highly complex equations and problems.
LLM’s on the other hand need to better adopt ‘computational thinking’ which involves defining and understanding a problem and and then breaking it down into pieces so that an LLM can calculate the answer, minimizing hallucinations and maximizing response accuracy.
*Toggle click to open benchmarks