# Steven Cao

ML engineer at Capable; ex-Cognition research engineer who trained SWE-1.6 and SWE-1.7; ex-Redwood Research; Harvard CS  
Capable  
San Francisco, California, United States

## At Cognition

- Role: Former Researcher (Research)
- Joined: Dec 2025
- How they joined: Directly
- Before Cognition: Research Intern, Redwood Research
- Now: Capable (https://www.linkedin.com/in/steven-duan-cao)
- Overlapped before Cognition with other Cognition people at: Harvard University (https://www.autumn.ai/resources/cognition/worked-together)

## Identity
- LinkedIn: https://www.linkedin.com/in/steven-duan-cao
- Domain: capable.com

## Notes
- Capable's Fall 2026 update describes him as having become "one of the ~five core research MLEs at Cognition, key for training Cognition's trillion-parameter-scale models", and says he brings "experience wrangling post-training large models, compute at scale, optimization of runs". (https://progress.capable.com/p/202610)
- Left Cognition in Jun 2026 and joined Capable (lab automation / longevity-biotech AI lab in Mission Bay, SF) as Machine Learning Engineer / MTS in Oct 2026. Capable announced the same month that it had raised a $38M seed (Cathay, Abstract, Fidji Simo, Andrej Karpathy and others). (https://progress.capable.com/p/202610)
- Co-author of 'SWE-1.7: Frontier Intelligence at a Fraction of the Cost' (the byline ends '... Steven Cao, Silas Alberti', with an equal-contribution marker). SWE-1.7 was RL-trained on a Kimi K2.7 base and scored 42.3 on FrontierCode 1.1, close to Claude Opus 4.7/4.8 and GPT-5.5 at a much lower cost per rollout. (https://cognition.com/blog/swe-1-7)
- Research Engineer at Cognition from Feb 2026 to Jun 2026; his LinkedIn role description reads, word for word: "trained swe 1.6 and 1.7 models". (https://www.linkedin.com/in/steven-duan-cao)
- Co-author of 'Introducing SWE 1.6: Improving Model UX' (Rohan Choudhury, Carlo Baronio, Ben Pan, Sam Lee, Eric Lu, Steven Cao, ...). SWE-1.6 was post-trained from scratch, shipped generally available in Windsurf at up to 950 tok/s, and came with a free tier for 3 months. (https://cognition.com/blog/swe-1-6)
- Listed as a co-author on Cognition's 'An Early Preview of SWE-1.6 and Research Update' post (by Carlo Baronio, Ben Pan, Sam Lee, Eric Lu, Steven Cao, Rohan Choudhury, Adam Zweiger, Ray ...). The preview improved SWE-Bench Pro by more than 10% over SWE-1.5. (https://cognition.com/blog/swe-1-6-preview)
- NVIDIA quoted him in its NVFP4 blog post: Cognition is seeing "significant latency and throughput gains" by using NVFP4 in large-scale reinforcement learning, "said Steven Cao, a member of Cognition's research team.". (https://developer.nvidia.com/blog/3-ways-nvfp4-accelerates-ai-training-and-inference/)
- His LinkedIn post on joining Capable, word for word: "Excited to be joining Capable to help push the limits of human biology! We have so much potential for longer, healthier, and more empowered lives." (https://www.linkedin.com/posts/steven-duan-cao_excited-to-be-joining-capable-to-help-push-activity-7514786099722887168-JbFd?utm_source=social_share_send&utm_medium=member_desktop_web&rcm=ACoAAHCanKIBY5mvNUmuesGnKpUAGt1HDtH6N_c)
- Started at Cognition as a Research Intern (Dec 2025 - Jan 2026, San Francisco) and converted to full-time Research Engineer within about two months. (https://www.linkedin.com/in/steven-duan-cao)
- Research Intern at Redwood Research (Apr - Jun 2025, SF), where he worked on "alignment faking evals". Capable describes this as work "on reducing alignment faking at Redwood Research". It was his first AI-safety lab role, during his first year at Harvard. (https://www.linkedin.com/in/steven-duan-cao)
- Studied Computer Science (Bachelor's) at Harvard University, with the LinkedIn listing 2024-2026. He was training frontier coding models at Cognition by age ~19-20 (his summary simply reads "20yo"). (https://www.linkedin.com/in/steven-duan-cao)
- Public headline says "ml @capable, prev @cognition, harvard". (https://www.linkedin.com/in/steven-duan-cao)
- Previous roles include Research Engineer at Cognition; Research Intern at Cognition; Research Intern at Redwood Research. (https://www.linkedin.com/in/steven-duan-cao)
- A newsletter tracking potential founders flagged him in September 2026 as 'Steven Cao in San Francisco (Potential Founder). Role: Researcher @ Cognition (<1 yr)', with Redwood Research as prior experience. (https://stealthmodefounders.com/p/9-7-26-top-founders-of-the-week)
- Location is San Francisco, California, United States. (https://www.linkedin.com/in/steven-duan-cao)
- Speaks Chinese and English (LinkedIn languages). (https://www.linkedin.com/in/steven-duan-cao)

## Experience
- Machine Learning Engineer at Capable (Oct 2026 - Present): lab automation
- Research Engineer at Cognition - San Francisco, California (Feb 2026 - Jun 2026): trained swe 1.6 and 1.7 models
- Research Intern at Cognition - San Francisco, California (Dec 2025 - Jan 2026)
- Research Intern at Redwood Research - San Francisco, California (Apr 2025 - Jun 2025): alignment faking evals

## Education
- Harvard University - Bachelor's, Computer Science - (Jan 2024 - Jan 2026)

## Sources
- https://www.linkedin.com/in/steven-duan-cao
- https://progress.capable.com/p/202610
- https://developer.nvidia.com/blog/3-ways-nvfp4-accelerates-ai-training-and-inference/
- https://cognition.com/blog/swe-1-7
- https://cognition.com/blog/swe-1-6
- https://cognition.com/blog/swe-1-6-preview
- https://www.linkedin.com/in/steven-duan-cao?miniProfileUrn=urn%3Ali%3Afsd_profile%3AACoAAD_ZmeUBkN_E3PNU9uO1-11WjXS4rhND2m8
- https://techcrunch.com/2025/07/14/cognition-maker-of-the-ai-coding-agent-devin-acquires-windsurf/
- https://emerginginvestigators.org/articles/22-288
- https://stealthmodefounders.com/p/9-7-26-top-founders-of-the-week
- https://ca.milesplit.com/meets/593943-605-league-cluster-1-2024/teams/17903
- https://www.linkedin.com/posts/steven-duan-cao_excited-to-be-joining-capable-to-help-push-activity-7514786099722887168-JbFd?utm_source=social_share_send&utm_medium=member_desktop_web&rcm=ACoAAHCanKIBY5mvNUmuesGnKpUAGt1HDtH6N_c


Researched with Autumn (https://www.autumn.ai) from public sources. Updated 2026-11-04. Profile: https://www.autumn.ai/resources/cognition/people/steven-cao · Everyone at Cognition: https://www.autumn.ai/resources/cognition/people