Claim Your Account
Lets get you verified!
Great! You may close this page
after you've sent us the code on ideascale.
You will receive a confirmation once your account is validated
To verify your ownership of this profile,
please send a personal message to
Lido Nation
on ideascale and include the code above.
Login to follow Justin
Justin
Claimed
Follow
Bio
Creating Hetzerk, a protocol for decentralized physical intelligence. and AI PhD student
Community Reviews across funding rounds
Justin Proposals (8)
Decentralized Physics Simulations
Solution: With SNET & Nunet we can run decentralized physics computations while rewarding computation, data, algorithms, and physics mining.
Decentralized Physics Tokenomics
Solution: To allow free decentralized physics computations we will integrate basic tokenomic features into our developing computational infrastructure
Decentralized AI Drug Development
Solution: Decentralized AI startup for computational drugs and therapeutics while rewarding computation & data, with percent of profits to community
Decentralized Physics Simulations
Solution: Free decentralized physics computations while rewarding computation & data for corporate, academic, and community involvement with Cardano
Hetzerk: A protocol for Decentralized Materials (Therapeutics at first) Discovery to bridge Large Industries, like Pharmaceuticals, to newer and better blockchain technologies leveraging Nunet
Solution: A protocol to facilitate major industry, academic, and scientific adoption of Cardano-based protocols for AI-assisted, privacy-sensitive physics computations in material and drug discovery
Truly Decentralized Large Language Model with Cardano SPOs
Solution: We will create a decentralized LLM by distributing its parameters across Cardano stake pool operators, integrated with Nunet, for enhanced security and resilience.
Zoda, Synthetic Data Protocol
Solution: Zoda, a protocol to harmonize disparate data sources into cohesive synthetic datasets, while community-curated will serve a wide range of applications, from AI training to scientific simulations.
Hetzerk: Decentralized Materials Discovery Platform
Solution: A protocol to facilitate major industry, academic, and scientific adoption of Cardano-based protocols for AI-assisted, privacy-sensitive physics computations in material and drug discovery
Monthly Reports
Error following
Successfully subscribed. Watch your inbox!
We will check for new reports and deliver them directly in your email.
Your email will not be shared with any 3rd party or be used for anything else.
More specific items can be located on the submitted document. Some highlights are device onboarding to Nunet, local installation of Nunet infrastructure, dock/nomad scripts for nunet deployment, and attending the Cardano Summit.
More specific items can be located on the submitted document. Some highlights are device onboarding to Nunet, local installation of Nunet infrastructure, dock/nomad scripts for nunet deployment, and attending the Cardano Summit.
Pytorch implementations are available with Nunet, we are figuring out right now how exactly to use them to integrate various computations and automated pipelines. Hard to tell at the current stage. We will finish setting up tests to examine how best to use the current framework and then develop a more concrete plan on how to incorporate everything Looking into incorporating proof-ilke algorithms in the calculation of physical properties, i.e. proof as computation for our physics mining applications [10% development / 0% testing] Looking to port over other open-source projects to integrate into a unified manner [05% development / 0% testing] We started development and initial testing of a protein/ligand complex generation algorithm that can sample different local configurations given amino acid sequences of the protein and ligand. [30% done] Training is overfitting, we need to somehow change the training data and/or architecture to improve results Generative model for thermodynamic quantities of single small molecules [100 development / 100 % testing] Generative model for thermodynamic quantities of multiple diverse small molecules [90 development / 70 % testing] The performance is not as good as for single models, we hope that integrating it into nunet can help clarify the global landscape and make training and performance faster/better. Pytorch implementations are available with Nunet, we are figuring out right now how exactly to use them to integrate various computations and automated pipelines. Hard to tell at the current stage. We will finish setting up tests to examine how best to use the current framework and then develop a more concrete plan on how to incorporate everything
Decentralized Physics Simulations September Report ‘constrained’ generative model of molecular configurations [90% development / 40% testing] Some bugs in the theory that were fixed, also numerical stabililty issues that are being worked on Energy/Force prediction from Molecular Dynamics trajectories [100% development / 100 % testing] Finished, development locally, next to integrate it into an automatable framework on nunet Bijective graph generative model of molecular compounds [100% development / 100% testing] Finished development locally, again, still need to integrate it into nunet Applied to Fit4Start competition / incubator in Luxembourg, results to come by end of september. If successful, we will be presenting at ArchSummit in Luxembourg in October Pytorch implimentations are available with Nunet, we are figuring out right now how exactly to use them to integrate various computations and automated pipelines. Hard to tell at current stage. We will finish setting up tests to examine how best to use the current framework and then develop a more concrete plan on how to incorporate everything Looking into incorporating proof-ilke algorithms in the calculation of physical properties, i.e. proof as computation for our physics mining applications [10% development / 0% testing] Doing some exploration tests with Julia Programming languages based on category theoretic models Looking to port over other open-source projects to integrate into a unified manner [05% development / 0% testing]
We are developing core algorithms right now that will be the basis for future growth, so things are looking good. Slower than hoped with Nunet, but so far, on schedule
We are developing core algorithms right now that will be the basis for future growth, so things are looking good. Slower than hoped with Nunet, but so far, on schedule
No
Our project is so far on track, and gaining a decent amount of traction through our friends at SingularityNet.
No
No
There seems to be a high level of interest in our project, especially in the SingularityNet community, whom have shown some interest in helping us develop our prototype.
We are happy to start the process.
No