Ai2

2 Jobs

312 Employees

About the Company

We are a Seattle-based non-profit AI research institute founded in 2014 by the late Paul Allen. We develop foundational AI research and innovation to deliver real-world impact through large-scale open models, data, robotics, conservation, and beyond.

Listed Jobs

Company Name: Ai2
Job Title: Research Internship, Advancing Open Agentic LLMs
Job Description: Job Title: Research Intern, Advancing Open Agentic LLMs Role Summary: Contribute to developing open-source agentic large language models (LLMs) capable of autonomous decision-making in dynamic environments. Focus on improving training, evaluation, and deployment of models that plan, reason, and interact using tools/APIs for real-world tasks. Expectations: 12-week full-time commitment; collaborative team-based research with publication opportunities; participation in projects like OLMo and Asta. Key Responsibilities: Build and optimize agentic LLM architectures; design benchmarks for multi-step reasoning and tool use; conduct experiments on training strategies; evaluate generalization, safety, and reliability of agentic systems. Required Skills: Strong background in AI/ML research; expertise in LLMs, reinforcement learning, or computational environments; programming skills in Python or related languages; ability to analyze complex datasets and synthesize insights. Required Education & Certifications: Pursuing PhD (or, in rare cases, MS/BS) in computer science, AI, or related field. Demonstrated research experience in LLM-based agents or agentic systems.

Seattle, United states

On site

25-01-2026

Company Name: Ai2
Job Title: Research Internship, FlexOlmo
Job Description: Job title: Research Intern – FlexOlmo Role Summary: 12‑week internship focused on designing, training, and evaluating large language models utilizing Mixture‑of‑Experts, long‑context capabilities, and retrieval methods. Interns lead a research project, build open‑source tools, and publish findings. Expectations: Deliver a fully documented research project, release trained models, produce a manuscript for a top‑tier conference, and actively collaborate with the research team. Key Responsibilities: - Design and execute experiments on MoE, LCLM, and retrieval architectures. - Implement and optimize training pipelines in PyTorch or equivalent. - Release models and code to the community with proper documentation. - Write and revise research papers for submission to NeurIPS, ICLR, ACL, EMNLP, or similar venues. - Attend weekly lab meetings, code reviews, and present progress updates. Required Skills: - Strong proficiency in machine learning and deep learning frameworks (PyTorch preferred). - Experience with large language models, training dynamics, scaling laws, and data curation. - Familiarity with Mixture‑of‑Experts, long‑context language models, or retrieval (preferred). - Ability to produce clean, reproducible code and maintain version control. - Excellent written and verbal communication; capacity to present technical results. Required Education & Certifications: - Current enrollment in a Ph.D., Master’s, or undergraduate program in Computer Science, Electrical Engineering, or a closely related field. - Demonstrated research experience in NLP, ML, or vision with a publication record in leading AI conferences (e.g., NeurIPS, ICLR, ICML, ACL, EMNLP, COLM).

Berkeley, United states

Remote

27-02-2026