Two-paper cap would cut ICML submissions by 45%, Miro Dudik estimates from last year's data
roydanroy · x · 2026-09-24
Based on last year's ICML submission data, Miro Dudik estimates that a cap of two submissions per author would reduce the number of papers by 45% — ignoring strategic effects like authors moving themselves to acknowledgements. Prominent researchers including Yarin Gal and Andrew Wilson joined the discussion, probing whether the analysis counts first authors only or also last authors.
More from Research
- VoiceArena launches Diarization Bench: 12 systems, NVIDIA Nemotron 3 leads at 14.72% DER — rohanpaul_ai · 2026-09-24
- Berkeley team trains 2.3B MoE Hybrid Mamba matching Llama-3.2-3B with <1% of pretraining FLOPs — berkeley_ai · 2026-09-24
- Why GPT-6's rumored recurrent-depth architecture could change inference economics — panic_in_the_galaxy · 2026-09-24
- An AUC 0.62 classifier cut 65,578 candidates to 145 DFT runs and found 6 new discoveries — bravo_abad · 2026-09-24
- Agentic Optimisation: what's fundamentally new for vision in the agent era — _krishna_murthy · 2026-09-24
- Google's WeatherNext 3 reads raw satellite data, cuts short-range temperature error up to 40% — anselm · 2026-09-24