Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BasedAI: A decentralized P2P network for Zero Knowledge Large Language Models (ZK-LLMs)

Sean Wellington

TL;DR

BasedAI tackles the privacy-performance trade-off in distributed AI by integrating Fully Homomorphic Encryption with LLMs in a decentralized Brains-based network. It introduces Cerberus Squeezing, a dynamic quantization and optimization framework, enabling ZK-LLMs to process encrypted prompts with reduced computational burden. The system couples incentive-aligned tokenomics, TFT-based incentive shaping, and governance (via Pepecoin) to sustain a scalable, privacy-preserving ecosystem. The work demonstrates a pathway toward practical privacy-preserving AI services across sectors such as healthcare and finance, while outlining mechanisms to mitigate centralization and ensure robust trust over time. The practical impact lies in enabling secure, censorship-resistant AI services that preserve data confidentiality without sacrificing performance in a decentralized setting.

Abstract

BasedAI is a distributed network of machines which introduces decentralized infrastructure capable of integrating Fully Homomorphic Encryption (FHE) with any large language model (LLM) connected to its network. The proposed framework embeds a default mechanism, called "Cerberus Squeezing", into the mining process which enables the transformation of a standard LLMs into encrypted zero-knowledge LLMs, or "ZK-LLMs", leveraging insights from generative adversarial networks for data privacy. This novel quantization mechanism empowers BasedAI miners to process and respond to prompts derived from User interaction with LLMs without the need for decrypting either the queries or their corresponding responses. The introduction of Cerberus Squeezing significantly improves performance degradation caused by quantized functions in current FHE-compliant computing environments by proactively optimizing calls between users, miners, and validators.

BasedAI: A decentralized P2P network for Zero Knowledge Large Language Models (ZK-LLMs)

TL;DR

BasedAI tackles the privacy-performance trade-off in distributed AI by integrating Fully Homomorphic Encryption with LLMs in a decentralized Brains-based network. It introduces Cerberus Squeezing, a dynamic quantization and optimization framework, enabling ZK-LLMs to process encrypted prompts with reduced computational burden. The system couples incentive-aligned tokenomics, TFT-based incentive shaping, and governance (via Pepecoin) to sustain a scalable, privacy-preserving ecosystem. The work demonstrates a pathway toward practical privacy-preserving AI services across sectors such as healthcare and finance, while outlining mechanisms to mitigate centralization and ensure robust trust over time. The practical impact lies in enabling secure, censorship-resistant AI services that preserve data confidentiality without sacrificing performance in a decentralized setting.

Abstract

BasedAI is a distributed network of machines which introduces decentralized infrastructure capable of integrating Fully Homomorphic Encryption (FHE) with any large language model (LLM) connected to its network. The proposed framework embeds a default mechanism, called "Cerberus Squeezing", into the mining process which enables the transformation of a standard LLMs into encrypted zero-knowledge LLMs, or "ZK-LLMs", leveraging insights from generative adversarial networks for data privacy. This novel quantization mechanism empowers BasedAI miners to process and respond to prompts derived from User interaction with LLMs without the need for decrypting either the queries or their corresponding responses. The introduction of Cerberus Squeezing significantly improves performance degradation caused by quantized functions in current FHE-compliant computing environments by proactively optimizing calls between users, miners, and validators.
Paper Structure (60 sections, 18 equations, 11 figures, 3 tables)

This paper contains 60 sections, 18 equations, 11 figures, 3 tables.

Table of Contents

  1. Introduction
  2. BasedAI: Decentralized Network Architecture
  3. Embedding Privacy at the Core
  4. The BasedAI Platform
  5. Participant Roles in the BasedAI Ecosystem
  6. Brain Owners
  7. Miners
  8. Validators
  9. Brain Permanent Memory
  10. Brain Dynamics and Operations
  11. Acquiring Brains
  12. Burn Method:
  13. Stake Method:
  14. Brain Deactivation Procedure
  15. Operating and Managing a Brain
  16. ...and 45 more sections

Figures (11)

  • Figure 1: A “Brain” from Based Labs
  • Figure 2: Pepecoin cost per Brain increases with each burn
  • Figure 3: Pepecoin supply potentially burned or staked in the Brain creation process
  • Figure 4: Comparison of Validator Incentives
  • Figure 5: Brains, Miners, and Validators
  • ...and 6 more figures