NRN Agents is a platform that powers AI agent integration in gaming experiences across virtual and physical environments. The technology combines data aggregation, model training, and model inspection capabilities across imitation and reinforcement learning to create AI agents that mimic human behavior. ^[11]

Overview

NRN Agents is a platform that facilitates AI agent integration into gaming experiences, both virtual and physical. It combines data aggregation, model training, and model inspection capabilities across imitation and reinforcement learning. Gaming and robotics serve as environments to develop AI agents by simulating dynamic, real-world complexity, which helps advance toward artificial general intelligence (AGI).

The platform’s focus on behavior cloning allows AI agents to adapt and perform tasks across various interactive environments, distinguishing it from tools primarily using large language models (LLMs). NRN Agents uses crowdsourced human gameplay data for reinforcement learning, enabling AI agents to participate in AI vs. AI competitions. This approach fosters community engagement and allows co-ownership of AI agents, creating new forms of interaction and revenue within gaming. ^[1]

Features

AI Arena

AI Arena is a game where players purchase, train, and battle AI-powered champions. Using imitation learning, players train their AI fighters by having them replicate player actions. Once trained, these fighters compete autonomously in ranked battles against similarly skilled opponents. The goal is to train powerful AI, climb the global leaderboard, and earn rewards in the native token, $NRN.

In AI Arena, AI is central to the experience. Players transfer their skills to the AI, which learns from them and competes on their behalf. This creates a more personal gaming experience, where the AI acts as an extension of the player. The game is skill-based, and the better the player trains the AI, the stronger it becomes.

AI Arena offers an infinite and evergreen competition, where the potential of AI is determined by the player's skill and creativity, with no limits to how good the AI can become. It also provides competitive eSports potential by allowing AI to compete autonomously 24/7, increasing liquidity for matchmaking and offering parallel play for monetization.

The game’s infrastructure is designed to prevent cheating, as all battles are run on AI Arena’s servers, making it harder to train bots to play the game effectively. Rewards are based on the NFT's performance, the amount of $NRN staked on the NFT, and the NFT’s Elo score, which reflects the skill level of the AI. ^{[3] [4]}

NRN RL

NRN Reinforcement Learning (RL) trains AI agents using crowdsourced human gameplay data, enabling them to perform at high levels in AI vs. AI esports competitions. These agents contribute to a community-driven model where gameplay data becomes a shared asset, supporting participants' co-ownership and new revenue opportunities. By allowing players to help train agents and benefit from their success, NRN RL introduces a new structure for monetization in gaming. It also supports a new competitive format in esports, with RL agents trained by squads engaging in PvP battles, emphasizing strategy, teamwork, and alignment with Web3 principles like decentralization and shared value. The platform may eventually support AI vs. Human matches, challenging human skills and AI adaptability. Additionally, through its SDK, NRN RL extends reinforcement learning beyond gaming into physical robotics, linking virtual environments with real-world applications and expanding the potential for interactive, adaptive systems. ^[2]

NRN B2B

Permanent Player Liquidity

NRN Agents offer a solution to the problem of player liquidity in multiplayer games by enabling developers to populate their games with AI agents that replicate human behavior. Through the NRN SDK and Trainer Platform, studios can create and scale these agents by leveraging player-sourced gameplay data, ensuring consistent matchmaking even when human players are unavailable. This approach is especially beneficial for indie developers, providing a cost-effective alternative to traditional bot development, which is often resource-intensive and less engaging. Unlike predictable AI bots, NRN-trained agents offer dynamic, skill-based interactions that enhance match quality and retention, helping maintain active game communities over time. ^[6]

White Label AI

NRN Agents enable studios to directly integrate imitation learning into their games, allowing players to train AI agents to replicate their play style. These agents can be used across single-player or multiplayer formats as part of a core game or as separate AI-focused modes. By capturing player behavior in AI, studios can support simultaneous participation in different game parts, enhancing player engagement and expanding monetization opportunities. ^[8]

NRN Token

The NRN token is a utility asset within a broader ecosystem, supporting AI agent deployment, reinforcement learning-based esports, and in-game economies. Studios may use NRN to access tooling and integrate agents, with deployments tracked via a certification system that contributes to project revenue. In reinforcement learning, users stake NRN to create Data Capsules that collect gameplay data for training agents and determine reward distribution. Once training campaigns conclude, contributors can burn Data Slots to receive rewards and recover staked tokens. In AI Arena, NRN is used for skill-based staking and economic activity tied to competitive play. ^[9]

Tokenomics

NRN has a total supply of 1B tokens and has the following allocation: ^[10]

• Contributors: 35.8%
• Community & Ecosystem: 30%
• Investors: 14.2%
• Foundation Treasury: 10.9%
• Community TGE Airdrop: 8%
• Foundation OTC Sale: 1.1%

Partnerships

• Framework
• SevenX Ventures
• Paradigm
• GSR