Roman Nekrasov
@romannekrasovaillm
Agentic mid-training, reinforcement learning with reward verification (RLVR), scaling agent environments, interleaved agent reasoning with tools
On GitHub since March 2025
1
Published Tools
1
Total Stars
0
Weekly Downloads
10
GitHub Followers
381
Public Repos
100/100
Avg Security
Published Tools
1 Agentacross 1 categoryromannekrasovaillm
A
Агентный TUI-харнесс на Rust под ML/RL-домен: 28 инструментов, 363 скилла, локальная Ариадна Qwen3.5-4B, 1 321 тест
Agentuncategorised
11 dir