Lego-RL
Collection
Harness-native RL for coding agents: the trained policy and the training task index. • 6 items • Updated • 11
AI & ML interests
We study code intelligence.
Recent Activity
View all activity
Papers
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents
SWE-Review: Closing the Loop on Issue Resolution with Agentic Code Review
ElvinDu518
updated a
collection about 11 hours ago
Mirror Lego-RL-2699 layout: add official source view, rename eval index, rewrite README
#2 opened about 12 hours ago
by 
ElvinDu518
Add SWE-bench Verified 500 as harbor task environments + index + README
#1 opened about 12 hours ago
by ElvinDu518
ElvinDu518
published a
dataset about 12 hours ago
YuxinJiang
updated 3
models about 13 hours ago
YuxinJiang
updated a
dataset about 21 hours ago
JierunChen
updated 3
collections 2 days ago
YuxinJiang
updated a
collection 2 days ago
YuxinJiang
published 2
models 2 days ago
ElvinDu518
updated a
model 2 days ago
yuantao428
authored a
paper 7 days ago
YuxinJiang
authored 3
papers 7 days ago
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents
Paper • 2608.17393 • Published • 24
SWE-Review: Closing the Loop on Issue Resolution with Agentic Code Review
Paper • 2607.06065 • Published • 11
What Makes Interaction Trajectories Effective for Training Terminal Agents?
Paper • 2606.03461 • Published • 3