{"version":"0.1","company":{"name":"YubHub","url":"https://yubhub.co","jobsUrl":"https://yubhub.co/jobs/skill/mechanistic-interpretability"},"x-facet":{"type":"skill","slug":"mechanistic-interpretability","display":"Mechanistic Interpretability","count":3},"x-feed-size-limit":100,"x-feed-sort":"enriched_at desc","x-feed-notice":"This feed contains at most 100 jobs (the most recently enriched). For the full corpus, use the paginated /stats/by-facet endpoint or /search.","x-generator":"yubhub-xml-generator","x-rights":"Free to redistribute with attribution: \"Data by YubHub (https://yubhub.co)\"","x-schema":"Each entry in `jobs` follows https://schema.org/JobPosting. YubHub-native raw fields carry `x-` prefix.","jobs":[{"@context":"https://schema.org","@type":"JobPosting","identifier":{"@type":"PropertyValue","name":"YubHub","value":"job_60da952d-d37"},"title":"Research Scientist, Interpretability","description":"<p><strong>About the role</strong></p>\n<p>When you see what modern language models are capable of, do you wonder, &quot;How do these things work? How can we trust them?&quot; The Interpretability team at Anthropic is working to reverse-engineer how trained models work because we believe that a mechanistic understanding is the most robust way to make advanced systems safe.</p>\n<p>We&#39;re looking for researchers and engineers to join our efforts. People mean many different things by &quot;interpretability&quot;. We&#39;re focused on mechanistic interpretability, which aims to discover how neural network parameters map to meaningful algorithms.</p>\n<p>A few places to learn more about our work and team at a high level are this introduction to Interpretability from our research lead, Chris Olah; a discussion of our work on the Hard Fork podcast produced by the New York Times, and this blog post (and accompanying video) sharing more about some of the engineering challenges we’d had to solve to get these results.</p>\n<p>Some of our team&#39;s notable publications include A Mathematical Framework for Transformer Circuits, In-context Learning and Induction Heads, Toy Models of Superposition, Scaling Monosemanticity, and our Circuits’ Methods and Biology papers.</p>\n<p>This work builds on ideas from members&#39; work prior to Anthropic such as the original circuits thread, Multimodal Neurons, Activation Atlases, and Building Blocks.</p>\n<p>We aim to create a solid foundation for mechanistically understanding neural networks and making them safe (see our vision post).</p>\n<p>In the short term, we have focused on resolving the issue of &quot;superposition&quot; (see Toy Models of Superposition, Superposition, Memorization, and Double Descent, and our May 2023 update), which causes the computational units of the models, like neurons and attention heads, to be individually uninterpretable, and on finding ways to decompose models into more interpretable components.</p>\n<p>Our subsequent work found millions of features in Sonnet, one of our production language models, represents progress in this direction.</p>\n<p>In our most recent work, we develop methods that allow us to build circuits using features and use this circuits to understand the mechanisms associated with a model&#39;s computation and study specific examples of multi-hop reasoning, planning, and chain-of-thought faithfulness on Haiku 3.5, one of our production models.</p>\n<p>This is a stepping stone towards our overall goal of mechanistically understanding neural networks.</p>\n<p>We often collaborate with teams across Anthropic, such as Alignment Science and Societal Impacts to use our work to make Anthropic’s models safer.</p>\n<p>We also have an Interpretability Architectures project that involves collaborating with Pretraining.</p>\n<p><strong>Responsibilities</strong></p>\n<ul>\n<li>Develop methods for understanding LLMs by reverse engineering algorithms learned in their weights</li>\n<li>Design and run robust experiments, both quickly in toy scenarios and at scale in large models</li>\n<li>Create and analyze new interpretability features and circuits to better understand how models work.</li>\n<li>Build infrastructure for running experiments and visualizing results</li>\n<li>Work with colleagues to communicate results internally and publicly</li>\n</ul>\n<p><strong>You may be a good fit if you:</strong></p>\n<ul>\n<li>Have a strong track record of scientific research (in any field), and have done some work on Interpretability</li>\n<li>Enjoy team science – working collaboratively to make big discoveries</li>\n<li>Are comfortable with messy experimental science. We&#39;re inventing the field as we work, and the first textbook is years away</li>\n<li>You view research and engineering as two sides of the same coin. Every team member writes code, designs and runs experiments, and interprets results</li>\n<li>You can clearly articulate and discuss the motivations behind your work, and teach us about what you&#39;ve learned. You like writing up and communicating your results, even when they&#39;re null</li>\n</ul>\n<p>To learn more about the skills we look for and how to prepare for this role, see our blog post – So You Want to Work in Mechanistic Interpretability?</p>\n<p>Familiarity with Python is required for this role.</p>\n<p><strong>Role Specific Location Policy:</strong></p>\n<ul>\n<li>This role is based in San Francisco office; however, we are open to considering exceptional candidates for remote work on a case-by-case basis.</li>\n</ul>\n<p>The annual compensation range for this role is listed below. For sales roles, the range provided is the role’s On Target Earnings (&quot;OTE&quot;) range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role.</p>\n<p>Annual Salary: $350,000-$850,000 USD</p>\n<p style=\"margin-top:24px;font-size:13px;color:#666;\">XML job scraping automation by <a href=\"https://yubhub.co\">YubHub</a></p>","url":"https://yubhub.co/jobs/job_60da952d-d37","directApply":true,"hiringOrganization":{"@type":"Organization","name":"Anthropic","sameAs":"https://www.anthropic.com/","logo":"https://logos.yubhub.co/anthropic.com.png"},"x-apply-url":"https://job-boards.greenhouse.io/anthropic/jobs/4980427008","x-work-arrangement":"hybrid","x-experience-level":"senior","x-job-type":"full-time","x-salary-range":"$350,000-$850,000 USD","x-skills-required":["Python","Mechanistic Interpretability","LLMs","Neural Networks","Circuits","Features","Model Computation"],"x-skills-preferred":[],"datePosted":"2026-04-18T15:44:56.628Z","jobLocation":{"@type":"Place","address":{"@type":"PostalAddress","addressLocality":"San Francisco, CA"}},"employmentType":"FULL_TIME","occupationalCategory":"Engineering","industry":"Technology","skills":"Python, Mechanistic Interpretability, LLMs, Neural Networks, Circuits, Features, Model Computation","baseSalary":{"@type":"MonetaryAmount","currency":"USD","value":{"@type":"QuantitativeValue","minValue":350000,"maxValue":850000,"unitText":"YEAR"}}},{"@context":"https://schema.org","@type":"JobPosting","identifier":{"@type":"PropertyValue","name":"YubHub","value":"job_20e650c2-d9c"},"title":"Research Scientist, Interpretability","description":"<p><strong>About the role:</strong></p>\n<p>When you see what modern language models are capable of, do you wonder, &#39;How do these things work? How can we trust them?&#39;</p>\n<p>The Interpretability team at Anthropic is working to reverse-engineer how trained models work because we believe that a mechanistic understanding is the most robust way to make advanced systems safe. We’re looking for researchers and engineers to join our efforts.</p>\n<p>People mean many different things by &#39;interpretability&#39;. We&#39;re focused on mechanistic interpretability, which aims to discover how neural network parameters map to meaningful algorithms. Some useful analogies might be to think of us as trying to do &#39;biology&#39; or &#39;neuroscience&#39; of neural networks using “microscopes” we build, or as treating neural networks as binary computer programs we&#39;re trying to &#39;reverse engineer&#39;.</p>\n<p><strong>Responsibilities:</strong></p>\n<ul>\n<li>Develop methods for understanding LLMs by reverse engineering algorithms learned in their weights</li>\n</ul>\n<ul>\n<li>Design and run robust experiments, both quickly in toy scenarios and at scale in large models</li>\n</ul>\n<ul>\n<li>Create and analyse new interpretability features and circuits to better understand how models work.</li>\n</ul>\n<ul>\n<li>Build infrastructure for running experiments and visualising results</li>\n</ul>\n<ul>\n<li>Work with colleagues to communicate results internally and publicly</li>\n</ul>\n<p><strong>You may be a good fit if you:</strong></p>\n<ul>\n<li>Have a strong track record of scientific research (in any field), and have done _some_ work on Interpretability</li>\n</ul>\n<ul>\n<li>Enjoy team science – working collaboratively to make big discoveries</li>\n</ul>\n<ul>\n<li>Are comfortable with messy experimental science. We&#39;re inventing the field as we work, and the first textbook is years away</li>\n</ul>\n<ul>\n<li>You view research and engineering as two sides of the same coin. Every team member writes code, designs and runs experiments, and interprets results</li>\n</ul>\n<ul>\n<li>You can clearly articulate and discuss the motivations behind your work, and teach us about what you&#39;ve learned. You like writing up and communicating your results, even when they&#39;re null</li>\n</ul>\n<p><strong>Role Specific Location Policy:</strong></p>\n<ul>\n<li>This role is based in San Francisco office; however, we are open to considering exceptional candidates for remote work on a case-by-case basis.</li>\n</ul>\n<p><strong>Logistics</strong></p>\n<p><strong>Education requirements:</strong> We require at least a Bachelor&#39;s degree in a related field or equivalent experience. <strong>Location-based hybrid policy:</strong> Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices.</p>\n<p><strong>Visa sponsorship:</strong> We do sponsor visas! However, we aren&#39;t able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this.</p>\n<p><strong>We encourage you to apply even if you do not believe you meet every single qualification.</strong> Not all strong candidates will meet every single qualification as listed.</p>\n<p style=\"margin-top:24px;font-size:13px;color:#666;\">XML job scraping automation by <a href=\"https://yubhub.co\">YubHub</a></p>","url":"https://yubhub.co/jobs/job_20e650c2-d9c","directApply":true,"hiringOrganization":{"@type":"Organization","name":"Anthropic","sameAs":"https://www.anthropic.com","logo":"https://logos.yubhub.co/anthropic.com.png"},"x-apply-url":"https://job-boards.greenhouse.io/anthropic/jobs/4980427008","x-work-arrangement":"hybrid","x-experience-level":"senior","x-job-type":"full-time","x-salary-range":"$350,000 - $850,000USD","x-skills-required":["Python","Mechanistic Interpretability","Neural Networks","Reverse Engineering","Experimental Science"],"x-skills-preferred":["Research","Engineering","Team Science","Communication"],"datePosted":"2026-03-08T13:48:39.765Z","jobLocation":{"@type":"Place","address":{"@type":"PostalAddress","addressLocality":"San Francisco"}},"employmentType":"FULL_TIME","occupationalCategory":"Engineering","industry":"Technology","skills":"Python, Mechanistic Interpretability, Neural Networks, Reverse Engineering, Experimental Science, Research, Engineering, Team Science, Communication","baseSalary":{"@type":"MonetaryAmount","currency":"USD","value":{"@type":"QuantitativeValue","minValue":350000,"maxValue":850000,"unitText":"YEAR"}}},{"@context":"https://schema.org","@type":"JobPosting","identifier":{"@type":"PropertyValue","name":"YubHub","value":"job_61280fe7-04a"},"title":"Researcher, Interpretability","description":"<p><strong>Job Posting</strong></p>\n<p><strong>Researcher, Interpretability</strong></p>\n<p><strong>Location</strong></p>\n<p>San Francisco</p>\n<p><strong>Employment Type</strong></p>\n<p>Full time</p>\n<p><strong>Location Type</strong></p>\n<p>Hybrid</p>\n<p><strong>Department</strong></p>\n<p>Safety Systems</p>\n<p><strong>Compensation</strong></p>\n<ul>\n<li>$295K – $445K • Offers Equity</li>\n</ul>\n<p>The base pay offered may vary depending on multiple individualized factors, including market location, job-related knowledge, skills, and experience. If the role is non-exempt, overtime pay will be provided consistent with applicable laws. In addition to the salary range listed above, total compensation also includes generous equity, performance-related bonus(es) for eligible employees, and the following benefits.</p>\n<ul>\n<li>Medical, dental, and vision insurance for you and your family, with employer contributions to Health Savings Accounts</li>\n</ul>\n<ul>\n<li>Pre-tax accounts for Health FSA, Dependent Care FSA, and commuter expenses (parking and transit)</li>\n</ul>\n<ul>\n<li>401(k) retirement plan with employer match</li>\n</ul>\n<ul>\n<li>Paid parental leave (up to 24 weeks for birth parents and 20 weeks for non-birthing parents), plus paid medical and caregiver leave (up to 8 weeks)</li>\n</ul>\n<ul>\n<li>Paid time off: flexible PTO for exempt employees and up to 15 days annually for non-exempt employees</li>\n</ul>\n<ul>\n<li>13+ paid company holidays, and multiple paid coordinated company office closures throughout the year for focus and recharge, plus paid sick or safe time (1 hour per 30 hours worked, or more, as required by applicable state or local law)</li>\n</ul>\n<ul>\n<li>Mental health and wellness support</li>\n</ul>\n<ul>\n<li>Employer-paid basic life and disability coverage</li>\n</ul>\n<ul>\n<li>Annual learning and development stipend to fuel your professional growth</li>\n</ul>\n<ul>\n<li>Daily meals in our offices, and meal delivery credits as eligible</li>\n</ul>\n<ul>\n<li>Relocation support for eligible employees</li>\n</ul>\n<ul>\n<li>Additional taxable fringe benefits, such as charitable donation matching and wellness stipends, may also be provided.</li>\n</ul>\n<p>More details about our benefits are available to candidates during the hiring process.</p>\n<p>This role is at-will and OpenAI reserves the right to modify base pay and other compensation components at any time based on individual performance, team or company results, or market conditions.</p>\n<p><strong>About the Team</strong></p>\n<p>The Interpretability team studies internal representations of deep learning models. We are interested in using representations to understand model behavior, and in engineering models to have more understandable representations. We are particularly interested in applying our understanding to ensure the safety of powerful AI systems. Our working style is collaborative and curiosity-driven.</p>\n<p><strong>About the Role</strong></p>\n<p>OpenAI is seeking a researcher passionate about understanding deep networks, with a strong background in engineering, quantitative reasoning, and the research process. You will develop and carry out a research plan in mechanistic interpretability, in close collaboration with a highly motivated team. You will play a critical role in helping OpenAI ensure future models remain safe even as they grow in capability. This will make a significant impact on our goal of building and deploying safe AGI.</p>\n<p>In this role, you will:</p>\n<ul>\n<li>Develop and publish research on techniques for understanding representations of deep networks.</li>\n</ul>\n<ul>\n<li>Engineer infrastructure for studying model internals at scale.</li>\n</ul>\n<ul>\n<li>Collaborate across teams to work on projects that OpenAI is uniquely suited to pursue.</li>\n</ul>\n<ul>\n<li>Guide research directions toward demonstrable usefulness and/or long-term scalability.</li>\n</ul>\n<p><strong>You might thrive in this role if you:</strong></p>\n<ul>\n<li>Are excited about OpenAI’s mission of ensuring AGI benefits all of humanity, and are aligned with OpenAI’s charter.</li>\n</ul>\n<ul>\n<li>Show enthusiasm for long-term AI safety, and have thought deeply about technical paths to safe AGI.</li>\n</ul>\n<ul>\n<li>Bring experience in the field of AI safety, mechanistic interpretability, or spiritually related disciplines.</li>\n</ul>\n<ul>\n<li>Hold a Ph.D. or have research experience in computer science, machine learning, or a related field.</li>\n</ul>\n<ul>\n<li>Thrive in environments involving large-scale AI systems, and are excited to make use of OpenAI’s unique resources in this area.</li>\n</ul>\n<ul>\n<li>Possess 2+ years of research engineering experience and proficiency in Python or similar languages.</li>\n</ul>\n<ul>\n<li>Are deeply curious.</li>\n</ul>\n<p><strong>About OpenAI</strong></p>\n<p>OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve our mission, we must encompass and value the many different perspectives, voices, and experiences that form the full spectrum of humanity.</p>\n<p style=\"margin-top:24px;font-size:13px;color:#666;\">XML job scraping automation by <a href=\"https://yubhub.co\">YubHub</a></p>","url":"https://yubhub.co/jobs/job_61280fe7-04a","directApply":true,"hiringOrganization":{"@type":"Organization","name":"OpenAI","sameAs":"https://jobs.ashbyhq.com","logo":"https://logos.yubhub.co/openai.com.png"},"x-apply-url":"https://jobs.ashbyhq.com/openai/c44268f1-717b-4da3-9943-2557f7d739f0","x-work-arrangement":"hybrid","x-experience-level":"senior","x-job-type":"full-time","x-salary-range":"$295K – $445K • Offers Equity","x-skills-required":["Python","Machine Learning","Deep Learning","Research Engineering","Computer Science"],"x-skills-preferred":["AI Safety","Mechanistic Interpretability","Quantitative Reasoning","Engineering"],"datePosted":"2026-03-06T18:39:59.202Z","jobLocation":{"@type":"Place","address":{"@type":"PostalAddress","addressLocality":"San Francisco"}},"employmentType":"FULL_TIME","occupationalCategory":"Engineering","industry":"Technology","skills":"Python, Machine Learning, Deep Learning, Research Engineering, Computer Science, AI Safety, Mechanistic Interpretability, Quantitative Reasoning, Engineering","baseSalary":{"@type":"MonetaryAmount","currency":"USD","value":{"@type":"QuantitativeValue","minValue":295000,"maxValue":445000,"unitText":"YEAR"}}}]}