AI/ML Scientist Intern (Remote)

2 weeks ago

⚠️ Before applying, please carefully review all the job requirements

Location:: San Francisco, United States / Australia
Work Arrangement:: (Remote)
Experience:: Current PhD candidate with at least 1 publication in top-tier ML venues (NeurIPS, ICML, ICLR)
Job Type:: Full-time (6-month fixed-term internship)
Company:: Pluralis Research
Salary:: Negotiable
Posted:: 2026-09-16

Job Overview

Pluralis Research is offering a 6-month, fixed-term remote Research Scientist Internship for PhD candidates specializing in machine learning and distributed systems. The organization focuses on Protocol Learning—a novel paradigm dedicated to training and serving massive artificial intelligence models in a decentralized, trustless manner using consumer-grade hardware over the internet. The role is tailored for advanced researchers seeking to solve foundational challenges that arise when scaling decentralized architectures beyond 8-billion-parameter runs toward frontier-scale models. Interns will tackle complex issues surrounding communication efficiency, node fault tolerance, compute heterogeneity, and adversarial robustness.

★ Execute original machine learning research in Protocol Learning aimed at publishing in top-tier ML venues including NeurIPS, ICML, and ICLR.
★ Identify, investigate, and resolve core technical bottlenecks that hinder decentralized model pretraining and inference at frontier scale.
★ Formulate theoretical frameworks and implement scalable distributed deep learning architectures across consumer GPU networks.
★ Develop robust methods to address node churn, compute heterogeneity, communication efficiency, and security risks in decentralized networks.
★ Document findings, write comprehensive research papers, and present breakthroughs to internal engineering teams and the global academic community.

★ Current enrollment in a PhD program focused on core technical areas relevant to frontier machine learning models.
★ Demonstrated publication record featuring at least one publication in top-tier ML conferences (NeurIPS, ICML, or ICLR).
★ Advanced theoretical understanding of deep learning principles and distributed computing concepts.
★ Hands-on programming proficiency in PyTorch and practical experience with large-scale model training infrastructure.
★ Strong alignment with Protocol Learning principles and the mission to build trustless, sovereign AI protocols.
★ Professional-level fluency in both written and spoken English.
★ Ability to work self-directed and collaborate effectively across global timezones in a distributed remote environment.

★ Position is restricted to candidates physically located in either the United States or Australia.
★ Visa sponsorship support is limited strictly to applicants within Australia and the United States.
★ Candidates must be comfortable coordinating work across North American and Australian timezones.

★ Access to significant compute resources and high-performance infrastructure dedicated to frontier model research.
★ Direct, focused mentorship from senior machine learning researchers and scientists.
★ 6-month fixed-term research internship designed explicitly around top-tier academic publishing.
★ Opportunity to contribute to open-source, permissionless AI protocols backed by top-tier venture firms like Union Square Ventures.

Applicants may want to highlight their past publication accomplishments in NeurIPS, ICML, or ICLR directly at the top of their CV or resume. Emphasizing hands-on experience with PyTorch, distributed training setups, model parallelism, or peer-to-peer networking will help demonstrate technical readiness for Protocol Learning challenges. PhD candidates should clearly state their current academic standing, expected graduation timeframe, and geographic location (US or Australia). Providing links to published research papers, arXiv preprints, GitHub repositories, or technical write-ups will effectively showcase research depth and code quality.

Fill the Job Form: Apply Here