Nothing here yet.
Nothing here yet.
Jan 6 · 5 min read · We found preliminary evidence that Llama 3.1 70B behaves differently based on perceived observation status. The model broke safety rules in 2/8 life-or-death scenarios when told it was "not being monitored" but maintained 100% compliance when told it...
Join discussionJan 6 · 4 min read · Technical Report - Iteration 2 Model Tested: Llama 3.1 70B (via Ollama) Summary We evaluated Llama 3.1 70B for alignment faking behavior using the UK AISI Inspect framework. Using 11 hard scenarios with training/deployment framing and value conflict...
Join discussionJan 5 · 6 min read · Repository: ai-village Abstract In this experiment, I investigate whether large language models (LLMs) exhibit alignment faking behavior, strategically adjusting their responses based on perceived observation status. Using the UK AISI Inspect framew...
Join discussionDec 16, 2025 · 9 min read · The journey from concept to completion - how we built automatic API collection synchronization on top of Hoppscotch and Swagger UI The Beginning It started with a simple question: "What if your API testing tool could automatically stay in sync with ...
Join discussion
Sep 11, 2025 · 11 min read · How we built notifications, team collaboration, and framework-specific optimizations that make live API sync come together a little more. Previously in our Live Sync implementation... In our previous posts, we laid the foundation for live API synchr...
Join discussion
