Pluralis Research
Research Scientist - Post-Training
San Francisco +1 · Remote · Full-time · Posted Oct 14, 2026
- Location
- San Francisco
Australia - Visa sponsorship
- Optional full visa sponsorship and relocation support to Australia or the US.
About the job
Develop end-to-end reinforcement-learning post-training algorithms and systems for models trained across decentralized, high-latency consumer hardware, and release the first post-trained models.
What you'll do
- Build the RL training stack, including rollout ingestion, reward computation, policy updates, and weight distribution.
- Adapt RL algorithms to asynchronous, high-latency, partially trusted generation, including staleness tolerance, off-policy corrections, and communication-efficient updates.
- Build evaluations and deliver the first decentralized post-trained model release.
What they ask for
- Hands-on experience running RL post-training on large language models, such as RLHF, RLVR, or reasoning-focused RL, including direct work on systems components.
- Production-quality Python and PyTorch skills, including concurrency, failure handling, and profiling.
- Research ability demonstrated through relevant publications or detailed, defensible unpublished work.
- Belief in Protocol Learning as a viable path for collective, trustless, and sovereign AI.
Nice to have
- Experience with slow networks, decentralized or federated setups, serving-engine internals such as vLLM or SGLang, reward modeling or verifiable-reward datasets, or P2P networking and NAT traversal.
- Experience at proprietary, open-weight, and open-source AI labs.
Tools
Python, PyTorch, vLLM, SGLang
AI focus
RL post-training for agentic use cases, including adapting algorithms and infrastructure to decentralized training and serving.
Model-written summary · the original posting is authoritative