Signal
Advertise
Signal
Advertise
Sign in
Discover trends that matter
Trending repositories
Daily
Weekly
Monthly
Yearly
Live mentions
Topics
GitHub trending
Repositories
Developers
Insights
Stats
Log in
johnhalloran321/mcp_safety_training — GitHub trending stats & insights | Trendshift
Featured
CrunchyCleaner
johnhalloran321/mcp_safety_training
#
NLP
DPO/SafeDPO/OPAD training + eval for teaching tool-using LLMs to refuse falsely-benign MCP exploits
Visit GitHub
Python
4
1 contributors
Apache License 2.0
website
Social mentions
Recent discussions about this repository across the web
6/RAG-Pref is the strongest result: ~3x refusal-rate improvement over baseline with no fine-tuning run required, ~3.7x combined with DPO/SafeDPO. Full implementation is in the repo now, not just the…
@convexDad · x.com
Repository activities
repository's daily and monthly activities across stars, forks, merged PRs, issues, and closed issues