Applied Researcher I (AI Foundations)

We need to output HTML fragment with sections: Overview, Team Description, The Ideal Candidate (maybe as Responsibilities? Actually the posting includes Overview, Team Description, In this role, you will: (list of duties), The Ideal Candidate (list of traits), Basic Qualifications, Preferred Qualifications, Salary info, Benefits, etc. We must use

for section headings (e.g., About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). We need to map content to appropriate headings. The instructions: "use

for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule)." So we should use those exact headings? It says e.g., meaning examples. We need to decide which sections are present. We should not repeat the job title as heading. We should not put Job ID etc. We need to include a section only when posting has real, specific content for it. So we need to decide which sections we have content for. We have: - Overview: description of Capital One AI vision. Could be "About the role" or maybe "Overview". But we need to pick from the list: About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. Overview could go under "About the role". That seems appropriate. Team Description and "In this role, you will:" could be "Responsibilities". We'll list duties. "The Ideal Candidate" could be "Qualifications" or "Skills". Probably "Qualifications" for basic and preferred qualifications. But we also have Basic Qualifications and Preferred Qualifications separate. Could combine under "Qualifications". Or we could have "Requirements" for basic qualifications and "Qualifications" for preferred? The list includes both Requirements and Qualifications. We could use Requirements for basic qualifications, Qualifications for preferred qualifications. Or we could combine. We also have Salary info -> Pay section. Benefits -> Benefits section. Schedule: not mentioned; we should not include Schedule if no content. We also have "Capital One will consider sponsoring a new qualified applicant for employment authorization" etc. That's maybe not needed. We need to preserve all substantive information; do not invent or summarize away detail. So we need to keep the details: bullet points etc. We need to use

• for any list of duties/requirements/benefits, and

  for prose. We must not include boilerplate like EEO, etc. Remove cookie/navigation/legal/EEO/equal-opportunity/scam-warning boilerplate, duplicated lines, and application instructions. Thus we need to strip out the EEO statements, accommodation contact, etc. Also remove "No agencies please." etc. We need to keep salary ranges, benefits statement, etc. We need to keep the part about "Capital One offers a comprehensive, competitive, and inclusive set of health, financial and other benefits that support your total well-being. Learn more at the Capital One Careers website. Eligibility varies based on full or part-time status, exempt or non-exempt status, and management level." That's benefits info. We also have "This role is also eligible to earn performance based incentive compensation, which may include cash bonus(es) and/or long term incentives (LTI). Incentives could be discretionary or non discretionary depending on the plan." That's also benefits/compensation. We need to keep that. Now we need to produce HTML fragment only, no or , no markdown fences. Let's structure:

  About the role

  Overview paragraph(s).

  Responsibilities

  • Partner with a cross-functional team ...
  • Leverage a broad stack of technologies ...
  • Build AI foundation models through all phases ...
  • Engage in high impact applied research ...
  • Flex your interpersonal skills to translate the complexity ...

  Requirements

  Basic Qualifications:

  • Currently has, or is in the process of obtaining, a PhD in Electrical Engineering, Computer Engineering, Computer Science, AI, Mathematics, or related fields, with an exception that required degree will be obtained on or before the scheduled start date or M.S. in Electrical Engineering, Computer Engineering, Computer Science, AI, Mathematics, or related fields plus 2 years of experience in Applied Research
  But we also have Preferred Qualifications list. Could be under

  Qualifications

  with sublist. But we need to avoid thin sections. Perhaps combine Basic and Preferred under Qualifications? But we have separate headings Requirements and Qualifications. We'll use Requirements for basic qualifications, Qualifications for preferred qualifications. Thus:

  Requirements

  Basic Qualifications:

  ...

  Qualifications

  Preferred Qualifications:

  ...
  We need to include all preferred qualifications items as list items. Preferred qualifications list is long with many bullet-like items separated by line breaks. We need to parse them. The text: "Preferred Qualifications PhD in Computer Science, Machine Learning, Computer Engineering, Applied Mathematics, Electrical Engineering or related fieldsLLMPhD focus on NLP or Masters with 5 years of industrial NLP research experienceMultiple publications on topics related to the pre-training of large language models (e.g. technical reports of pre-trained LLMs, SSL techniques, model pre-training optimization)Member of team that has trained a large language model from scratch (10B + parameters, 500B+ tokens) Publications in deep learning theory Publications at ACL, NAACL and EMNLP, Neurips, ICML or ICLROptimization (Training & Inference)PhD focused on topics related to optimizing training of very large deep learning models Multiple years of experience and/or publications on one of the following topics: Model Sparsification, Quantization, Training Parallelism/Partitioning Design, Gradient Checkpointing, Model Compression Experience optimizing training for a 10B+ model Deep knowledge of deep learning algorithmic and/or optimizer design Experience with compiler designFinetuningPhD focused on topics related to guiding LLMs with further tasks (Supervised Finetuning, Instruction-Tuning, Dialogue-Finetuning, Parameter Tuning) Demonstrated knowledge of principles of transfer learning, model adaptation and model guidanceExperience deploying a fine-tuned large language model" It seems there is missing punctuation and spaces. We need to preserve the content as is, but we can format as list items splitting by logical sections. However we must not invent or summarize away detail. We need to keep the exact text? The instruction: "Preserve ALL substantive information; do NOT invent or summarize away detail." So we should not change the meaning, but we can reformat into list items as long as we keep all info. Splitting into multiple list items is okay as long as we don't lose info. However we must be careful not to inadvertently omit any detail. We could keep the entire preferred qualifications as a single paragraph (

  ) but the rule says use

  • for any list of duties/requirements/benefits. Preferred qualifications is a list, so we should use
    • . We need to split into sensible items. We can split by the phrases separated by line breaks in the original (though they are not clearly separated). Let's try to identify each bullet: - PhD in Computer Science, Machine Learning, Computer Engineering, Applied Mathematics, Electrical Engineering or related fields - LLMPhD focus on NLP or Masters with 5 years of industrial NLP research experience - Multiple publications on topics related to the pre-training of large language models (e.g. technical reports of pre-trained LLMs, SSL techniques, model pre-training optimization) - Member of team that has trained a large language model from scratch (10B + parameters, 500B+ tokens) - Publications in deep learning theory - Publications at ACL, NAACL and EMNLP, Neurips, ICML or ICLR - Optimization (Training & Inference) - PhD focused on topics related to optimizing training of very large deep learning models - Multiple years of experience and/or publications on one of the following topics: Model Sparsification, Quantization, Training Parallelism/Partitioning Design, Gradient Checkpointing, Model Compression - Experience optimizing training for a 10B+ model - Deep knowledge of deep learning algorithmic and/or optimizer design - Experience with compiler design - Finetuning - PhD focused on topics related to guiding LLMs with further tasks (Supervised Finetuning, Instruction-Tuning, Dialogue-Finetuning, Parameter Tuning) - Demonstrated knowledge of principles of transfer learning, model adaptation and model guidance - Experience deploying a fine-tuned large language model We need to preserve the exact wording, but we can add punctuation for readability? Might be considered altering. Safer to keep the text as is but split at logical points. However the original lacks spaces between some items (e.g., "fieldsLLMPhD"). That's a typo; we should preserve it? The instruction says preserve ALL substantive information; do not invent or summarize away detail. Changing "fieldsLLMPhD" to "fields. LLMPhD" would be adding a period and space, which is inventing? It's adding punctuation not present. Might be considered altering. Safer to keep the exact string as is, but we could still make it a single list item containing that exact string. However that would be a huge blob and not a list. But the rule says use
      • for any list of duties/requirements/benefits. If we make one
      • containing the whole block, that's still a list (single item). That satisfies the rule. It preserves all info exactly. So we could do:

        Qualifications

        Preferred Qualifications:

        • PhD in Computer Science, Machine Learning, Computer Engineering, Applied Mathematics, Electrical Engineering or related fieldsLLMPhD focus on NLP or Masters with 5 years of industrial NLP research experienceMultiple publications on topics related to the pre-training of large language models (e.g. technical reports of pre-trained LLMs, SSL techniques, model pre-training optimization)Member of team that has trained a large language model from scratch (10B + parameters, 500B+ tokens) Publications in deep learning theory Publications at ACL, NAACL and EMNLP, Neurips, ICML or ICLROptimization (Training & Inference)PhD focused on topics related to optimizing training of very large deep learning models Multiple years of experience and/or publications on one of the following topics: Model Sparsification, Quantization, Training Parallelism/Partitioning Design, Gradient Checkpointing, Model Compression Experience optimizing training for a 10B+ model Deep knowledge of deep learning algorithmic and/or optimizer design Experience with compiler designFinetuningPhD focused on topics related to guiding LLMs with further tasks (Supervised Finetuning, Instruction-Tuning, Dialogue-Finetuning, Parameter Tuning) Demonstrated knowledge of principles of transfer learning, model adaptation and model guidanceExperience deploying a fine-tuned large language model
        That preserves the exact string. Similarly for Basic Qualifications we could keep as a single
      • . Now Responsibilities: we have a list of duties after "In this role, you will:" with bullet-like sentences separated by line breaks. We'll make each a
      • . Now Benefits: we have a paragraph about benefits and incentive compensation. Could be a
        • for each benefit? But it's prose. We could make a

          for the benefits description. The rule says use

          • for any list of duties/requirements/benefits. Benefits is not a list but a description; we can use

            . However there is also a sentence about "Capital One offers a comprehensive, competitive, and inclusive set of health, financial and other benefits that support your total well-being. Learn