Maksim Ivanov

github linkedin

research

Mitigating Artifact Drift in Agent Benchmark Generation 2026

Introduces ERP-Bench.

Terminal-Bench 3 2026

Authored one of 100 accepted benchmark tasks.

Can Coding Agents Be General Agents? 2025

ICLR 2026 blogpost submission.

writing

Minding the Data Quality Gap

More essays on Substack.

art by @poetengineer__

Thinking in the notational niche