FULL STORY

Oxford's Yarin Gal Slams LLM-Written Papers; arXiv Responds

Oxford's Yarin Gal warned that LLM-written papers are flooding ML research with homogenized prose and urged arXiv to ban them. arXiv said it would detect but not block such submissions, and the debate over AI in scientific writing continues.

2026-09-22 ~ 2026-09-23 · 4 episodes · 11 posts

Episode 1 · Scholars debate LLM-written papers as ML papers grow homogeneous (2026-09-22, 4 posts)

Oxford's Yarin Gal warned that ML papers increasingly show homogeneous, hollow LLM-style writing, proposing policies requiring authors to write themselves, and predicted most LLM-written papers won't stand the test of time — a claim questioned by Thomas Dietterich.

Episode 2 · Oxford's Yarin Gal: LLMs are poor at real science without scaffolding (2026-09-23, 3 posts)

Oxford's Yarin Gal argues LLMs are poor at science without scaffolding: in his tests, an AI research assistant collapsed open-ended hypotheses into known work and misread papers, requiring human correction.

Episode 3 · arXiv Responds to LLM Paper Controversy: Detect but Don't Block (2026-09-23, 2 posts)

Oxford's Yarin Gal proposed that arXiv ban LLM-written papers detected via watermarks, but arXiv responded that it will flag such content for stricter review rather than reject it outright.

Episode 4 · Oxford's Yarin Gal: Irreproducible LLM Experiments Are Just Bad Experiments (2026-09-23, 2 posts)

Oxford's Yarin Gal argued that LLM-written experiments that fail to replicate are no different from any other irreproducible experiments, and suggested building an arXiv-like platform for AI-generated research.