The title, by tech reports I don't mean arXiv submissions, but reports of a large model, like say Kimi K3, DeepSeek, Gemini, Mistral Leanstral, etc. Is it much above, above, much below, below or equal to a first author A* paper?