“Ellie Bacausch of Prime Intellect stated that Anthropic's Mythos model is intentionally designed to perform poorly on tasks related to frontier LLM research, which she views as a negative development for the research community.”