Observed arrival · 2026-11-10
Irregularities: a replication reference for deep learning
A research reference tracks reported deep-learning effects and gathers studies that test where they hold, weaken, or fail to reproduce.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Readers checking deep-learning research claims
- Worth noticing
- Evidence groups are not a score: the page says studies can appear in multiple groups, so their counts overlap.
Field notes
An entry divides a reported effect into lettered claims, including whether it exists, the settings it needs, and how it works. The guide assigns suggested statuses ranging from a single-paper report to independent reproduction, contested, or unclear evidence. It also cautions that studies can fall into multiple evidence groups, so group counts overlap rather than forming a score.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue