Open AI training and model evaluation roles that call for LLM Research expertise, gathered from 5 platforms. Part of Data & Machine Learning.
Pay figures cover the 25 open jobs that publish an hourly rate in USD. They are published rates, not a guarantee of income. Last checked on October 8, 2026.
This role involves conducting security testing and red-teaming exercises on AI systems to identify vulnerabilities, weaknesses, and failure modes. The position is designed for security professionals who can think like attackers and use creative methods to stress-test AI models and evaluate their safety, bias, and policy compliance.
This role involves evaluating and improving large language models' understanding of physics by designing advanced problems, authoring ground-truth solutions, and auditing AI-generated reasoning across quantum mechanics, electrodynamics, thermodynamics, and other domains. It is suited for PhD-level applied physicists who want to contribute to AI training while working independently on a flexible schedule.
This contract role asks mathematicians to turn dense informal proofs into structured, machine-checkable Lean 4 formalizations. The work also involves helping AI researchers understand where automated provers fail and how to improve formal verification pipelines. It is aimed at people with strong proof-writing backgrounds who want to work at the intersection of mathematics and AI.
This contract role involves turning rigorous informal mathematical proofs into machine-checkable Lean 4 formalizations and identifying gaps and hidden assumptions along the way. It is aimed at mathematicians with graduate-level training who enjoy precise proof writing and formal verification. The work is part-time and remote, with collaboration alongside AI researchers.
This is a remote, hourly contract role in which experienced machine learning and software engineers write step-by-step reasoning logs and training datasets that teach frontier AI models how to solve complex technical problems. The work also involves auditing and refining AI-generated reasoning chains for errors and logical gaps. It is aimed at technical professionals who can work independently and on their own schedule.
This remote, asynchronous contract role asks holders of a Master's or PhD in a STEM or highly technical field to build hard, multi-disciplinary problems and review AI-generated solutions to strengthen how AI systems reason. It suits self-directed researchers with publications or substantial original projects who can explain complex ideas precisely in writing. Hours are flexible and the work may extend into further projects.
This freelance role asks mathematicians to design demanding problems and reference solutions that test how AI systems reason. The work also includes checking model outputs for accuracy and giving structured feedback, with benchmarks spanning undergraduate to PhD-level material. It suits people with advanced mathematical training who can work independently and remotely.
This remote contract role asks mathematicians to write and audit formal proofs in Lean 4, contributing to large formal libraries such as mathlib. It suits people with an advanced degree in mathematics who can turn informal reasoning into machine-checkable proofs and want to help train mathematically capable language models.
This remote contract role asks mathematicians to write, check and audit formal proofs in Lean 4, feeding large formal libraries and helping train language models that reason mathematically. It is aimed at people with advanced training in mathematics who can work independently and asynchronously.
This remote contract role asks mathematicians to write and check formal proofs in Lean 4, strengthening large formal libraries and the mathematical reasoning of language models. It suits people with a graduate degree in mathematics who are comfortable working independently and asynchronously.
This remote contract role asks mathematicians to translate informal theorems and arguments into machine-checkable proofs in Lean 4, helping build the formal foundations that advanced AI systems rely on. It is aimed at candidates with graduate training in mathematics or a closely related field who enjoy rigorous proof and can work independently on an asynchronous schedule.
This is a remote, freelance contract for mathematicians who write formal proofs in Lean 4. The work involves turning informal mathematical arguments into machine-checkable proofs and contributing to large formal libraries, while supporting the training of mathematically capable language models. It suits people with a graduate background in mathematics who are comfortable working independently and on their own schedule.
This contract role asks mathematicians to formalize advanced theorems in Lean 4 and to contribute to large formal libraries such as mathlib. It also covers auditing existing formal proofs and helping train mathematically capable language models. It suits holders of a graduate degree in mathematics who can work independently and asynchronously on flexible hours.
This contract role asks mathematicians to formalize advanced theorems and proofs in Lean 4 and to audit existing formal proofs for correctness. The work also helps train language models to reason mathematically and feeds into large formal libraries such as mathlib. It suits people with graduate training in mathematics who are comfortable working independently and asynchronously.
This freelance role asks mathematicians to turn informal proofs into machine-checkable formal proofs in Lean 4, helping shape how advanced AI systems reason. The work covers theorems from many mathematical fields, contributions to large formal libraries, and auditing existing proofs. It suits holders of advanced mathematics degrees who can work independently on flexible, remote schedules.
This remote contract asks a physics expert to write advanced problems that test how well AI models reason, along with clear step-by-step solutions. The work also includes checking AI-generated answers for accuracy and quality of reasoning, and helping researchers refine benchmarks. It is aimed at people holding a master's degree in physics or a related field, and no prior AI experience is needed.
This contract role asks engineers to write and review step-by-step reasoning traces that show how a language model should plan, use tools, and reach decisions. The output becomes training data meant to make AI systems more reliable on complex, real-world tasks. It suits experienced machine learning engineers and researchers who can work remotely and on their own schedule.
This freelance role asks senior machine learning practitioners to write step-by-step reasoning traces that record how an expert plans, uses tools, and decides through hard technical problems, so that large language models learn to reason the same way. It suits people with deep ML or AI systems backgrounds who can explain their thinking clearly and work independently on open-ended tasks.
This remote contract role involves building coding benchmarks and data pipelines that test how well frontier AI models reason about, debug, and write software. It is aimed at senior software engineers with several years of professional experience who can work across large, multi-language codebases.
Expert ML research for evaluating advanced deep learning work (generative models, reinforcement learning, transformer architectures).
Referral link to micro1's job list: search for this role there. Open this exact job
Add your resume once. Every open job is scored against your profile, and we email you only when a strong match appears. Free.
FAQ
31 LLM Research AI training jobs are open on SideHustler today. 1 was posted in the last 7 days. Listings are rechecked against the official postings: the latest check on this list was on October 8, 2026.
Among the 25 open roles that publish an hourly rate in USD, pay runs from $15 to $200 per hour, with a median of $110. These are the rates the platforms publish, not a guarantee of income or hours.
Alignerr (19), Mercor (8), micro1 (2), Handshake AI (1) and Vetto (1). You apply on the platform itself, which handles screening, contracts and payment.
The most requested areas of expertise on these listings are LLM Research, AI Evaluation, Mathematics, Machine Learning, Deep Learning. Each listing states its own requirements.
Apply now opens the official posting on the platform, where you complete the application. You need no account on SideHustler. To see which roles fit your background first, add your resume and every open job is scored against your profile.
