Tag
A curated thread covering three notable AI papers: MiniMax Sparse Attention for efficient long-context inference, Self-Harness for self-improving agent scaffolds, and Agents' Last Exam benchmark for measuring agent economic value.
This paper introduces Self-Harness, a new paradigm where LLM-based agents iteratively improve their own operating harness—prompts, tools, and control flow—without human engineers or stronger external agents, achieving significant performance gains across multiple models.