Researcher Lets Claude Review His ICLR Paper's Reviews, and It Roasts the Community's Favorite Assumption
rao2z · x · 2026-11-14
Researcher rao2z shared the peer reviews for one of his ICLR 2027 papers and had Claude comment on them. Claude's take: the reviews came out about as well as one could hope for a paper whose job is to tell the community its favorite assumption is unjustified. A lighthearted look at AI-conference peer review and the awkward position of work that challenges mainstream assumptions.
More from Fun
- Post-training engineer jokes he's "going out of context" from pretraining advances — auto_grad_ · 2026-11-14
- Recreating Sunshine's observation room in Three.js with WebGL — nptacek · 2026-11-14
- Syntherrarium: a playable three.js instrument that grows tiny worlds as you play — nptacek · 2026-11-14
- Flying 15,000m over a rainforest in a browser tab: Three.js' 'metaverse' moment — jradoff · 2026-11-14
- OpenAI shows off limited-edition ModRetro Chromatic handheld in black collector's box — OpenAIDevs · 2026-11-14
- "Slopcore" AI music video built with Claude, Suno, Runway and Flow — Yamapama · 2026-11-14