Economist working in the field of macroeconomics, International Finance and Trade, Economic Theory, and Productivity
How media typically covers Naman Jain
Based on 4 scored articles
Referenced in coverage
“Author of research on reward hacking in LLM agents and the Reward Hacking Benchmark.”
“Developed R2E-Gym, performed data filtering for RL datasets, designed DeepSWE agent scaffold, and trained verifiers for test-time scaling.”
“Wrote about Cursor's hybrid online-offline eval process for model quality assessment.”
Cursor's custom semantic search embedding model improves coding agent accuracy by 12.5% on average and increases code retention by 2.6% on large codebases compared to traditional grep-based search.
“Co-authored technical content about semantic search implementation in Cursor's coding agent.”