Ethereum Foundation Announces ‘Deep Funding Christmas’ Mini-Contest, Offering $20,000 In Rewards

Non-profit organization supporting Ethereum, Ethereum Foundation announced the launch of the Deep Funding Christmas mini-contest. The initiative aims to create a distilled human judgment mechanism to predict the historical funding amounts received by open-source projects. The contest is open now and will run until 11:59:59 AoE on January 26th, with results announced on January 27th. 

Currently, the contest dataset includes 3,410 comparisons across 245 repositories, with 2,387 available for training and 1,023 for scoring. Participants are tasked with making 1,023 predictions about the relative funding received between any two projects, which will be tested against actual data. New submissions, such as open-source models, valuable datasets, or infrastructure that benefits other participants, are eligible for up to $20,000 in prizes. 

What Is Deep Funding And How Does Mini Contest Work?

The Deep Funding initiative operates with three interconnected components. It begins with an unweighted dependency graph that identifies all contributions forming part of a valuable project. This is followed by submissions of deep learning models that assign weights to the relative value of each contribution. Finally, a human judgment mechanism performs manual comparisons of a subset of contributions to select the model that aligns most closely with human preferences. The winning model is then used to allocate funding based on its proposed weights.

This technology is designed to support all relevant dependencies, extending beyond those that apply for funding in a specific round. Its objectives include eliminating the need for extensive public campaigning or cultivating favor with decision-makers to secure funding. Additionally, it facilitates the operation of more frequent funding rounds for a larger number of projects by replicating expensive funding mechanisms, ensuring accessibility and fairness on a broad scale.

Do you want your AI agent to be more than a Twitter reply guy?

Are you a ML modeller looking for a challenge?

Well, here's your chance at some glory!

Introducing the Deep Funding Christmas special: a mini contest to predict funding to open source dependencies!

🧵 below pic.twitter.com/V09cNZhcrT

— Devansh Mehta (@TheDevanshMehta) December 25, 2024

According to the contest guidelines, developers are encouraged to create a model capable of predicting relative funding amounts for 1,023 comparisons across 245 Ethereum open-source repositories. The provided dataset includes funding data from Open Collective, Gitcoin, and Optimism retroactive rounds for repositories that participated in these rounds. This data is structured as relative comparisons based on decisions made by actual funders, resulting in 3,410 total comparisons.

Participants will later use their models in the main contest, beginning at the end of January, to evaluate 4,303 repositories. The objective is to approximate how much funding all relevant repositories would have received if they had participated in similar funding rounds. However, the focus in the main contest is not on the funding data itself but on predicting each repository’s relative importance as ranked by open-source maintainers and a jury, with funding data serving only as a proxy.

This mini-contest serves as a trial run for a larger contest set to launch at the end of January. In the main contest, participants will apply their models to determine weights for a dependency graph of 4,303 Ethereum repositories and key dependencies, including consensus clients, developer tools, and programming languages. Winners in the main contest will allocate $170,000 to the 4,303 key Ethereum repositories based on predicted weights, with an additional $60,000 allocated for infrastructure that supports contest participants.

The post Ethereum Foundation Announces ‘Deep Funding Christmas’ Mini-Contest, Offering $20,000 In Rewards appeared first on Metaverse Post.