QueensGambit
Joined 8 years ago
Karma
924
Submissions
345
Member since
May 7
Recent submissions
Comment · 19 days ago
I got $10K GCP bill one month because of ClaudeBot. Its likely some users were trying to replicate our software by using Claude code + chrome plugin. If I was spending $10K on traffic, I can only imagine the money cloners must be spending on tokens.
Comment · 8 months ago
Factorization is key here. It separates dataset-level structure from observation-level computation so the model doesn't waste capacity rediscovering structure.
I've been arguing the same for code generation. LLMs flatten parse trees into token sequences, then burn compute reconstructing hierarchy as hidden states. Graph transformers could be a good solution for both: https://manidoraisamy.com/ai-mother-tongue.html
Comment · 9 months ago
― Epictetus