Karl Cobbe is an OpenAI researcher associated with the Procgen Benchmark for measuring RL generalization, with verifier-based approaches to math word problems (GSM8K), and with foundational contributions to OpenAI o1. OpenAI and academic publications document these lines of work on sample efficiency, generalization, and solution verification.123
At a glance
- Current role
- Researcher
- Company
- OpenAI
- Sources cited
- 3
- Profile reviewed
Accomplishments
Procgen and verifiers
Cobbe coauthored the Procgen Benchmark paper and OpenAI's Procgen release for procedurally generated RL environments. Separate work on training verifiers to solve math word problems introduced GSM8K-style verification for selecting correct solutions.12
o1
OpenAI's o1 contributions page lists Cobbe among foundational contributors to the reasoning model effort.3
Timeline
Where they work
- OpenAI 102 people in the wiki
Related people
- Isa FulfordAlso at OpenAI
- Shixiang Shane GuAlso at OpenAI
- Bob McGrewAlso at OpenAI
- Shengjia ZhaoAlso at OpenAI
- Ilya KostrikovAlso at OpenAI
- Ethan PerezEvaluation and benchmarks · Language models
- Owain EvansLanguage models · Evaluation and benchmarks
- Akshita BhagiaLanguage models · Evaluation and benchmarks
- Marina DanilevskyEvaluation and benchmarks · Language models
- Samuel R. BowmanEvaluation and benchmarks · Language models
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
3 sources · profile reviewed Oct 24, 2026 · data updated Oct 24, 2026