Hugging Face
Models
Datasets
Spaces
Buckets
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
Datasets:
ninty-seven
/
stepguard_data
Like
1
Modalities:
Text
Formats:
json
Languages:
English
Size:
1K - 10K
ArXiv:
arxiv:
2608.24777
Tags:
agent-safety
tool-use
step-level-safety
supervised-fine-tuning
reinforcement-learning
Libraries:
Datasets
pandas
Polars
+ 1
License:
apache-2.0
Dataset card
Data Studio
Files
Files and versions
xet
Community
1
main
stepguard_data
/
rl4k
Commit History
Oct 2
28 days ago
09:00
Remove metadata field from RL training examples
ninty-seven
Verified signature
aadeae0
Remove the metadata column from all 4,000 RL examples and update the dataset card. Keep sample order and all remaining fields unchanged.
Sep 22
14:31
Use concise StepGuard RL sample identifiers
ninty-seven
Verified signature
e9e84ae
Replace 4,000 verbose top-level sample IDs with stepguard-rl-000001 through stepguard-rl-004000. Preserve row order and all other sample fields. Document the identifier format in the dataset card.
Jul 28
14:45
Add rl4k/agentguard_rl4k_grpo.jsonl
ninty-seven
Verified signature
4a2b032
