Spaces:

lucid987654
/

code-review-env-v3

Sleeping

App Files Files Community

code-review-env-v3

7.08 MB

Ctrl+K

Ctrl+K

2 contributors

History: 38 commits

Kinchi

add ablation experiments: truncation baseline + tag removal analysis

1619d0b about 2 months ago

code_review_env
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
data
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
grpo_output
add ablation experiments: truncation baseline + tag removal analysis about 2 months ago
scripts
add ablation experiments: truncation baseline + tag removal analysis about 2 months ago
server
🔐 CodeReviewEnv v3: Agentic Security Investigation with Thinking Budget about 2 months ago
tests
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
.gitattributes

1.57 kB
add: real training curves from 200-step GRPO run + improvement panel about 2 months ago
.gitignore

358 Bytes
clean: submission-ready repo about 2 months ago
ENV.md

8.72 kB
fix: broken file path references in JUDGES.md and ENV.md about 2 months ago
JUDGES.md

11.6 kB
fix: broken file path references in JUDGES.md and ENV.md about 2 months ago
PAPER.md

19.1 kB
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
README.md

9.14 kB
add ablation experiments: truncation baseline + tag removal analysis about 2 months ago
SAFEGUARDS.md

9.46 kB
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
app.py

50.9 kB
fix: handle Gradio SSR zero-input calls in run_demo about 2 months ago
blog_post.md

13.1 kB
add ablation experiments: truncation baseline + tag removal analysis about 2 months ago
demo.py

8.12 kB
🔐 CodeReviewEnv v3: Agentic Security Investigation with Thinking Budget about 2 months ago
eval_baseline.py

10.3 kB
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
inference.py

5.73 kB
🔐 CodeReviewEnv v3: Agentic Security Investigation with Thinking Budget about 2 months ago
metacognitive_reward.py

11.5 kB
v2.2: switch to Qwen3-1.7B + unlock full lengths + real calibration logging about 2 months ago
openenv.yaml

10.2 kB
clean: submission-ready repo about 2 months ago
requirements.txt

183 Bytes
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
rubrics.py

8.69 kB
clean: submission-ready repo about 2 months ago
run.sh

3.23 kB
v2: calibrated metacognition as RL + inference-time budget + transfer eval about 2 months ago
train_colab.ipynb

11 kB
fix: Red Team tab SSR crash + incremental training curves + storytelling polish about 2 months ago
train_grpo.py

53.9 kB
tune: LR 5e-6, warmup 0.05, KL 0.04 + EMA smoothing + trend line about 2 months ago
train_sft_warmup.py

7.02 kB
fix: SFT max_seq_length→max_length + 200 episodes (not 150) about 2 months ago
transfer_eval.py

10.5 kB
v2: calibrated metacognition as RL + inference-time budget + transfer eval about 2 months ago