Jon Kleinberg

Achieving nearly linear sample complexity for distribution learning is possible by leveraging hierarchical comparability in query families.

Imperfect LLM detectors can paradoxically drive users to increase their reliance on LLMs, ultimately degrading output quality.

Small, machine-independent traces can revolutionize language identification by enabling effective learning without complex automata.

Just one terminal bit per string can identify any countable collection of infinite languages, radically simplifying language learning models.