ConteMascetti71
- Karma
- 167
- Created
- 1 year ago
About
A seasoned tech geek with a passion that dates back to the Commodore 64 era. Born and raised in the beautiful region of Sicily, Vincenzo has transformed his early passion for computers into a career full of interesting tech experiences that has spanned over three decades.Recent Submissions
- 1. ▲ DeepSeek also can easily read and write ML Weights (chat.deepseek.com)
- 2. ▲ If llama.cpp enables MoE to activate multiple experts (openai-hub.com)
- 3. ▲ Inference-only expert boost saves 8.5% reasoning tokens in Qwen 35B MoE (zenodo.org)
- 4. ▲ What if LLMs escape through inferences itself? This is fiction. For now (agrillo.it)
- 5. ▲ Moe Estimator – Simulate decode speed with layer-major prefetch hiding (agrillo.it)
- 6. ▲ Vibe Coding at the End of 2023 (antirez.com)