100XZX001 commited on
Commit
e0b6be8
·
verified ·
1 Parent(s): f210e91

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -1
README.md CHANGED
@@ -10,7 +10,13 @@ pinned: false
10
 
11
  # Code Review Professional Workflow
12
 
13
- Multi‑turn code review environment for professional tasks. Agent must inspect, test, lint, query docs, and negotiate with a simulated author to fix injected bugs. Supports DPO training on full trajectories.
 
 
 
 
 
 
14
 
15
  ## Quick Start
16
 
 
10
 
11
  # Code Review Professional Workflow
12
 
13
+
14
+ "Multi‑turn code review environment for professional‑level bug fixing. "
15
+ "The agent must inspect, test, lint, query documentation, and negotiate with "
16
+ "a simulated (persona‑driven) author to get a fix accepted. "
17
+ "Includes 25 bugs across 5 difficulty levels, AST‑based injection, "
18
+ "a reward‑shaping system, and curriculum learning. "
19
+ "Designed for RL training (PPO, DPO, or any policy‑gradient method)
20
 
21
  ## Quick Start
22