ToolGrad flips tool-use data generation from prompts to solutions
Most tool-use datasets start with a hypothetical user request, then spend expensive agentic search finding a valid API path. ToolGrad reverses that order: it builds a ground-truth tool chain first, then uses textual-gradient feedback to turn the solution into a query. Google reports that the approach produces more complex, long-horizon examples at lower cost, while models trained on them match proprietary systems on tests involving unseen tools.