Weighted mean baseline "exploits" virtual cell challenge, Stanford professor warns
anshulkundaje · x · 2026-11-20
Stanford professor Anshul Kundaje escalates his critique of the virtual cell challenge: after noting the mean baseline is formidable, he points to its "older and stronger sibling", the weighted mean baseline. While acknowledging the winning team played by the rules, he argues this hyper-optimized, biologically unfounded exploit is not the path to virtual cells — the challenge design confounds data advantage with methodological superiority.
More from Research
- MIT's Hiroshi Ishii shares Tangible Media Group's Creative Tension principles — ishii_mit · 2026-11-20
- Train a real LLM entirely in your browser with this interactive demo — morgymcg · 2026-11-20
- Ex-OpenAI Researcher: English Is the New ML Programming Language, Lean the Type System — lukaszkaiser · 2026-11-20
- S1CogBench probes System One decision models across 24 cognitive tasks — Dengzhe Hou · 2026-11-20
- AFFECT from KAIST makes world models dynamics-faithful via action effect matching — John Won · 2026-11-20
- Mind2Act World: a 180-task benchmark exposing robots' mind-action composition failures — Mind2Act-World · 2026-11-20