Hugging Face daily papers·1d agoTokenCast: Forecasting Token Consumption During LLM Agent Execution#tokencast#llm-agents#token-forecasting 2 sources