17 Nov
|
Nameless
|
Canada
You need to enable Java Script to run this app. body { overflow: hidden; } .center { align-items: center; display: flex; flex-direction: column; height: 100vh; justify-content: center; width: 100vw; } @keyframes fade-in { from { opacity: 0; } to { opacity: 1; } } .fade-in { animation-delay: 750ms; animation-direction: normal; animation-duration: 500ms; animation-fill-mode: forwards; animation-iteration-count: 1; animation-name: fade-in; animation-timing-function: ease-in-out; opacity: 0; } @keyframes spin { from { transform: rotate(0deg); } to { transform: rotate(360deg); } } .spinner { animation-duration: var(--transition Duration Slow, 600ms); animation-iteration-count: infinite; animation-name: spin; animation-timing-function: linear; border-color: var(--color Neutral600Transparent, rgba(125, 134, 153, 0.25)); border-radius: 100%; border-style: solid; border-top-color: rgba(125, 134, 153, 0.9); border-width: var(--border Width Heavy, 3px); display: block; height: var(--spacing Xxlarge, 48px); width: var(--spacing Xxlarge, 48px); } window.__app Data = {"dd RumApplication Id":"80e0bf43-e772-45ac-858b-1e6dc0f1f415","dd RumClient Token":"pub1a87c7036063ee7d4b7914b397e6324e","environment":"production","maintenance Mode":false,"scheduling Maintenance Mode":false,"organization":{"organization Id":"951a1fca-ef74-47dd-9932-d2e6ae54beb2","name":"Cohere","public Website":"https://cohere.com/","custom Jobs Page Url":null,"hosted Jobs Page Slug":"cohere","allow JobPost Indexing":true,"is Demo Org":false,"theme":{"colors":{"version":"1","color Primary600":"#355146","color Primary900":"#D9A6E5"},"show JobFilters":true,"show Location Address":false,"show Teams":true,"show Autofill Applications Box":true,"logo Wordmark Image Url":"https://app.ashbyhq.com/api/images/org-theme-wordmark/951a1fca-ef74-47dd-9932-d2e6ae54beb2/c275a67b-c32a-426c-bc5a-e64c488df273/3c9efb6c-4f13-431e-a109-127485109bfe.png","logo Square Image Url":"https://app.ashbyhq.com/api/images/org-theme-logo/951a1fca-ef74-47dd-9932-d2e6ae54beb2/c275a67b-c32a-426c-bc5a-e64c488df273/f9bad60a-761a-4d26-9d99-39bda6bbaed4.png","application Submitted Success Message":"Your application has been submitted! Thank you for your interest in a career growth opportunity at Cohere!","job Board TopDescription Html":null,"job Board Bottom Description Html":null,"job Posting Back Url":null,"__typename":"Theme"},"app Confirmation Tracking Pixel Html":null,"recruiting Privacy Policy Url":"https://cohere.com/applicant-privacy","active Feature Flags":["AiInterviewer","AiInterviewer Updated Time Estimates","AiNotetaker Candidate Consent Policies","Approval Notes From Email","Booking Link Reschedule Limits","Canonicalize Schools","Core Multiple Brands","Demo OrgBanners","Education History Form Field","Explicit Automated Processing OptOut Confirmation","Job Board Cookie Consent Controls","Job Board Etag Validation","Job Posting Application Deadlines","Job Posting ListV2","Job Posting Workplace Type Override Secondary Locations Too","Linked InIntegration Require Company Id","Linked InTOSUpdate","Location External Name","Migrate Google Recaptcha Automated Legal","Reject Base64Encoded Resumes Front End","Reject Base64Encoded Resumes Throw Error Front End","Reject Invalid File Type OnUpload","Reject Urls InCandidate Name","Scheduling Direct Booking LogDirect Booking Events Json","SchedulingGCal Send Updates ForCancel Events","Scheduling Reschedule AndCancellation Reasons NonUser","Scheduling UseLocale InTime Formatting","Show Counts OnJob Board Filter Options","Social Sharing Html Meta Settings","Use Google Recaptcha Universal Key","Validate File Type OnUpload"],"timezone":"America/New_York","candidate Schedule Cancellation Reason Requirement Status":"Off","show Cookie Banner":false,"__typename":"Organization"},"posting":{"id":"8a73f59d-41b7-42ed-8cc8-99afa0a2ba5f","title":"Member of Technical Staff - Post-Training (Reinforcement Learning & Generalization)","department Name":"Modeling","department External Name":"Modeling","location Name":"Europe","location Address":null,"workplace Type":"Remote","employment Type":"Full Time","description Html":"Who are we?
Cohere is the leading security-first enterprise AI company. We build cutting-edge foundation AI models and end-to-end products that are designed to solve real-world business problems.
Weâre training and deploying frontier models for enterprises who are building AI systems. We believe that our work is instrumental to the widespread adoption of AI and we are looking for folks that want to be part of that.
We obsess over what we build. Each one of us is responsible for contributing to increasing the capabilities of our models and the value they drive for our customers. Cohere is a team of researchers, engineers, designers, and more, who are all passionate about their craft.
We are a global technology company headquartered in Toronto with key offices in London, New York City, San Francisco, Montreal, Paris, Berlin and Seoul. Join us!
Key Overview:
As a Member of Technical Staff in Post-Training, you will train our large language models (LLMs) with reinforcement learning and develop methods that help them generalize beyond the distribution they were trained on. You will design the training recipes, environments,
and evaluations that tell us whether a model has truly learned a skill or has only fitted its training tasks, and turn what works into the recipes behind our production models. You will work closely with the post-training data infrastructure, evaluation, and pre-training teams.
Key Responsibilities
- Design, run, and analyse large-scale RL post-training experiments (e.g., RLHF, RL from verifiable rewards).
- Research and develop methods that improve out-of-distribution generalization, such as reward design, environment and task diversity, curricula, and regularization.
- Build held-out evaluations that measure generalization to unseen tasks, domains, and formats, and that catch reward hacking and benchmark overfitting.
- Diagnose training dynamics and failure modes, including instability, reward over-optimization, and capability regressions.
- Turn successful research into robust, reproducible training recipes for production models.
Qualifications
- Strong research and software engineering skills, with hands-on experience training LLMs.
- Experience applying reinforcement learning to LLMs (e.g., PPO, GRPO, RLHF, RLVR), ideally at scale.
- Track record of work on generalization, robustness, or distribution shift, through publications or shipped models.
- Rigorous experimental practice: clean ablations, careful controls, and sound conclusions from noisy results.
- Experience with distributed training (e.g., JAX, PyTorch, FSDP, Megatron) is a plus.
- Excellent communication skills.
- Deep understanding of data quality principles and best practices: demonstrates willingness to dive deep into data while possessing strong intuition for assessing the quality of datasets and anticipating how different data characteristics will impact model training outcomes.
Working Location
This role can be based remotely or from one of our office locations listed on the job description - there is no minimum in-office qualification requirement. We care most about hiring exceptional people regardless of locations, though please check the location listed on the posting for guidance around the core time zone or working hours alignment expected for the role.
Full time Employees at Cohere enjoy these Perks:
- A weekly lunch stipend of $75/£75 or equivalent in your local currency for lunch.
- Full health and dental benefits, including a separate budget for mental health.
- RRSP matching, 401K, Pension Scheme.
- 100% Parental Leave top-up for up to 6 months, for either parent.
- Annual enrichment benefits:
Arts & culture, fitness/wellness, quality time, and a workspace improvement credit.
Education & learning stipend for conferences, courses, and coaching.
- 6 weeks of paid vacation (30 working days!)
- Budget for traveling to other offices if you are remote, plus an annual company offsite.
How and Where We Work:
- Cohere is remote-friendly, but we also have offices in Toronto, London, New York City, San Francisco, Montreal, Paris, Berlin and Seoul with more opening soon.
- For those in the office: a daily lunch program, plenty of snacks, and regular community and social events.
- For those not near an office: a co-working benefit so you can work alongside others in your city.
- Everyone receives a $500 home office stipend to set up your workspace properly.
If any of the above doesnât line up exactly with your experience, we still encourage you to apply.
We strive to create an inclusive work environment for all; we welcome applicants from all backgrounds and are committed to providing equal opportunities. Should you require any accommodations during the recruitment process, please submit an Accommodations Request Form, and we will work together to meet your needs.
We may use AI-enabled tools to screen and assess applicants against the criteria for this position. This helps our recruiters identify potentially qualified candidates, but it doesn't limit the applications our recruiters may review or consider.
Beware of Scams: Cohere will never ask for payment or third-party services (e.g., CV writing) as part of our hiring process. All legitimate roles are listed on the Cohere careers page and Linked In only, with all communications from Cohere employees coming from an @cohere.com or @cw.cohere email alias. If jobs are viewed on other sites then please verify these through our official careers page.
","is Listed":true,"is Confidential":false,"team Names":["Modeling"],"secondary Location Names":["New York","Toronto"],"compensation Tier Summary":"CA$295K â CA$535K ⢠Offers Equity ⢠Multiple Ranges","compensation Tiers":[{"id":"0b3d1fb5-fa5a-40c2-affc-b493831b5d89","title":"Canada - Toronto (GTA), Montreal","tier Summary":"Base Salary CA$295K â CA$535K ⢠Offers Equity","__typename":"Job Posting Compensation Tiers"},{"id":"7d239aea-6513-4ebb-9004-b3449d92d066","title":"USA - New York,
California and Wasghington States","tier Summary":"Base Salary $205K â $380K ⢠Offers Equity","__typename":"Job Posting Compensation Tiers"},{"id":"ea904ffe-0da0-45fb-97c5-4a01133c6753","title":"USA - All Other States","tier Summary":"Base Salary $175K â $325K ⢠Offers Equity","__typename":"Job Posting Compensation Tiers"}],"application Deadline":null,"compensation Tier Guide Url":null,"scrapeable Compensation Salary Summary":"CA$295K - CA$535K","compensation Philosophy Html":"
Cohere is committed to fair and transparent pay practices. The compensation range listed for this role reflects the expected compensation. Actual compensation offered will be determined by factors such as location, level, job-related knowledge, skills, education, and experience.
","application Limit Callout Html":"
Please Note: we have set up limits for applications. The following limits apply to applications for all jobs:
- Candidates may not apply more than 5 times in any 90 day span
- Candidates may not re-apply to the same role within 365 days if not presented with an offer
","should AskFor Texting Consent":false,"should AskFor Whats AppConsent":false,"candidate Texting Privacy Policy Url":null,"candidate Texting Terms AndConditions Url":null,"legal Entity Name ForTexting Consent":"Cohere","automated Processing Legal Notice":null,"__typename":"Job Posting Details"},"job Board":null,"router Prefix":"/","recaptcha Universal Public Site Key":"6LezdY0tAAAAAEnollNLCAI0z1VDGwM7AzGrU4XZ"}; // A Vite manifest key: the entry's HTML filename relative to the build root. // Single-entry apps leave entry Html unset. const ENTRY_KEY = "index.html"; fetch("https://cdn.ashbyprd.com/frontend_non_user/446c70668dc93ffef6f135bdaa62e3deb46a64df/.vite/manifest.json").then(function (res) { return res.json() }).then(function (manifest) { const entry Data = manifest[ENTRY_KEY]; if (entry Data == null) { throw new Error("Vite manifest is missing entry: " + ENTRY_KEY); } // Walk the entry's static imports transitively. Both the CSS and the JS a // page needs can sit on shared chunks rather than on the entry record, and // the manifest lists only each chunk's direct imports -- so reading the // entry alone silently drops stylesheets and preloads. Depth-first // post-order puts a dependency's CSS ahead of its importer's, which is the // cascade order Vite uses in the HTML it generates for the same entry. const chunks = []; const visited = {}; function collect Chunks(key) { if (visited[key] === true) { return; } visited[key] = true; const chunk = manifest[key]; if (chunk == null) { return; } (chunk.imports ?? []).for Each(collect Chunks); chunks.push(chunk); } collect Chunks(ENTRY_KEY); // css[] holds output paths, not digests, but every manifest record carries // an integrity for its own file -- so a stylesheet Vite emitted as its own // record can be matched by path. The entry's CSS has no record and loads // unhashed. const integrity ByFile = {}; Object.keys(manifest).for Each(function (key) { const record = manifest[key]; if (record != null && record.file != null && record.integrity != null) { integrity ByFile[record.file] = record.integrity; } }); const sheets = []; const seen Sheets = {}; chunks.for Each(function (chunk) { (chunk.css ?? []).for Each(function (sheet) { if (seen Sheets[sheet] === true) { return; } seen Sheets[sheet] = true; sheets.push(sheet); }); }); let bundle Loaded = false; function load Bundle() { if (bundle Loaded === true) { return; } const el = document.create Element("script"); el.set Attribute("type", "module"); el.set Attribute("crossorigin", ""); el.set Attribute("integrity", ); el.set Attribute("src", "https://cdn.ashbyprd.com/frontend_non_user/446c70668dc93ffef6f135bdaa62e3deb46a64df/" + ); document.head.append Child(el); bundle Loaded = true; } chunks.for Each(function (chunk) { const preload = document.create Element("link"); preload.rel = "modulepreload"; preload.href = "https://cdn.ashbyprd.com/frontend_non_user/446c70668dc93ffef6f135bdaa62e3deb46a64df/" + chunk.file; // Carry the SRI hash on the preload too: Chrome/Firefox reuse the // preloaded bytes for the integrity-bearing script tag, so an integrity-less // preload silently bypasses the SRI check. Without it here, SRI is never // enforced on the code-split chunks (the bulk of the app), in any browser. preload.set Attribute("integrity", chunk.integrity); preload.set Attribute("crossorigin", ""); document.head.append Child(preload); }); if (sheets.length > 0) { const loaded Sheets = []; sheets.for Each(function (sheet) { const link = document.create Element("link"); link.rel = "stylesheet"; link.type = "text/css"; link.href = "https://cdn.ashbyprd.com/frontend_non_user/446c70668dc93ffef6f135bdaa62e3deb46a64df/" + sheet; link.media = "all"; const sheet Integrity = integrity ByFile[sheet]; if (sheet Integrity != null) { link.set Attribute("integrity", sheet Integrity); // SRI on a cross-origin stylesheet requires CORS mode; the CDN sends // access-control-allow-origin: *. link.set Attribute("crossorigin", ""); } link.onload = function () { loaded Sheets.push(sheet); if (loaded Sheets.length === sheets.length) { load Bundle(); } }; link.onerror = load Bundle; document.head.insert Before(link, document.get Element ById("vite-preload")); }); } else { load Bundle(); } });
📌 Member of Technical Staff - Post-Training (Reinforcement Learning & Generalization) (Canada)
🏢 Nameless
📍 Canada