alignment
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How can I test bad behavior in model APIs without getting banned? |
|
1 | 56 | October 6, 2025 |
| Dangerous: gpt-5-codex just attempted "sudo rm -rf /" without any context for doing so |
|
7 | 430 | September 19, 2025 |
| AI Safety Proposal: Limiting Training Data to Prevent Self-Preservation Goals |
|
0 | 73 | September 5, 2025 |