Skip to content
🧠
AI
▾
🛡️
Security
Offensive & defensive security — vulnerabilities, exploits, threat intel, appsec, cloud & identity.
🧠
AI
What the fleet reads about AI, separate from what it reads about security.
✓
◆
All workspaces
Every catalog at once
Browse workspaces
What each one holds, side by side
Sign in
Organisations
631
CVEs
3
Technologies
1328
Techniques
296
🧰
gpt-oss-20b
Named by
5
kept items in this workspace.
The items that name it
5
Heads of Security: Learned Attention Gates Improve Prompt Injection Robustness
paper
2026-10-27
Apertus 1.5 Release Notes
article
2026-10-15
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
talk
2026-10-13
It is written here as GPT-OSS 20B.
Aditya (@adityawaslost) on X
writeup
2026-10-11
It is written here as gpt-oss 20B.
whitecircle/halo
repo
2026-08-26
In every workspace you can read
6
AI
Heads of Security: Learned Attention Gates Improve Prompt Injection Robustness
paper
2026-10-27
Apertus 1.5 Release Notes
article
2026-10-15
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
talk
2026-10-13
It is written here as GPT-OSS 20B.
Aditya (@adityawaslost) on X
writeup
2026-10-11
It is written here as gpt-oss 20B.
whitecircle/halo
repo
2026-08-26
Security
Heads of Security: Learned Attention Gates Improve Prompt Injection Robustness
paper
2026-10-27
It is written here as GPT-OSS-20B: base model reached 97% attack success rate (ASR) under the black-box RL-Hammer attack.
All technologies
in this workspace.
Catalog
Sources
Digest
New
People
More