Research Scientist – Post-Training

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

📋 Description

• Develop the end-to-end RL post-training stack, which includes rollout ingestion, reward computation, policy updates, and sending updated weights back to the network.

• Establish the vision for the post-training stack and oversee its development.

• Modify RL algorithms for asynchronous, high-latency, and partially trusted generation scenarios.

• Tackle challenges related to staleness tolerance, off-policy corrections, and efficient policy updates through communication.

• Create evaluations to showcase improvements in the model.

• Achieve the delivery of the first decentralized post-trained model from execution to public artifact.

⛳️ Requirements

• Practical experience in executing RL post-training on large language models, including RLHF, RLVR, or reasoning-focused RL.

• Familiarity with rollout generation, asynchronous training loops, and weight synchronization.

• Proficient in Python and PyTorch at a production-quality level.

• Experience with concurrency, failure management, and profiling prior to optimization.

• Published work in RL post-training, asynchronous or distributed RL, or related areas, or unpublished work that can be defended in detail.

• A strong belief in Protocol Learning as a feasible method for achieving collective, trustless, and sovereign AI.

• Excellent command of English, both written and spoken, at a professional level.

• Ability to work effectively across different time zones.

• Experience in training over slow networks or with decentralized or federated configurations is advantageous.

• Knowledge of vLLM or SGLang serving internals is a plus.

• Background in reward modeling or working with verifiable-reward datasets is a bonus.

• Familiarity with P2P networking and NAT traversal is beneficial.

• Experience in proprietary, open-weight, and open-source AI labs is an advantage.

🏝️ Benefits

• Significant equity ownership alongside a competitive base salary.

• Flexible working environment.

• Comprehensive visa sponsorship and relocation assistance to either Australia or the US for exceptional candidates.

• Opportunity to engage with open research challenges and contribute to some of the first solutions.