Research Scientist – Decentralized Training and Inference Verification

This is a fully remote position, open to applicants in United States, +2 more locations.

📋 Description

• Identify and maintain the threat model addressing risks posed by malicious or negligent employees during pre-training, post-training, and inference phases.

• Evaluate potential threats such as training disruptions, denial-of-service attacks, free-riding, model poisoning, backdoors, data extraction, reputation manipulation, and reward manipulation.

• Develop statistical verification methodologies with defined error rates.

• Optimize verification tests using stringent benchmarks.

• Manage false positives and false negatives across diverse hardware configurations, including various GPUs and Macs.

• Create and operate the verification service within the inference pathway.

• Confirm the authenticity of tokens originating from the specified model and sampling parameters.

• Validate that training contributions are accurate as claimed.

⛳️ Requirements

• Developed a calibrated statistical decision-making system with established error rates and effectively managed its limitations.

• Extensive knowledge in statistics and probability.

• Proficient in designing experiments, calibrating decision thresholds, and justifying claimed error rates.

• Understanding of verifying untrusted computing environments, including statistical testing, re-execution methods, cryptographic proofs, and trusted hardware.

• Capability to evaluate which verification strategies are suitable for a permissionless network.

• Commitment to Protocol Learning as a feasible third path for creating collective, trustless, and sovereign AI.

• Acquainted with large-scale pre-training and reinforcement learning post-training methodologies.

• Knowledgeable about decentralized machine learning security and adversarial threat models, such as poisoning, Sybil attacks, collusion, and replay attacks.

• Experience in proprietary, open-weight, and open-source AI laboratories.

• Proficient in professional-level English, both written and spoken.

🏝️ Benefits

• Attractive equity options for key technical contributors, alongside a competitive base salary.

• Flexible work environment with a globally distributed team.

• Optional full visa sponsorship and relocation assistance to Australia or the US.