adam
for ML engineers, optimizer researchers
Why this scored 9
every term, weightedIts fastest-moving item, measured against the pace of its own source
Whether that velocity is itself speeding up, as a per-hour rate
How many independent communities its own items come from
Decays to zero over 14 days, counted from when we first saw it
Subtracted once something is big and old — sized by its biggest item, aged from when we first saw it
Weights are hand-tuned, not learned — we're calibrating them against realized trends as history accumulates. On an entity's first sighting there's no previous reading to compare against, so acceleration starts from a neutral prior rather than a measurement, and velocity falls back to engagement over its whole lifetime until a second reading exists. Full methodology
Signal history
7-day window (free)The evidence
The live items this entity's score aggregates — every community independently talking about it right now. This is the corroboration, shown, not claimed.
- 119
A Statistical Viewpoint on Adam, Shampoo and Muon
Statistical comparison of Adam, Shampoo, and Muon optimizers. · for ML engineers, optimizer researchers
arxivSteadyai17d ago - 219
Second-Moment Memory in Coordinatewise Adam
Analyzes second-moment memory in coordinatewise Adam, impacting optimizer tuning. · for ML engineers
arxivSteadyai80d ago - 319
Spectral Allocation: Why Muon Outperforms Adam, and How to Improve Muon
Muon optimizer shows better convergence than Adam, suggesting a switch for faster model training. · for ml engineers
arxivSteadyai71d ago - 418
Error Feedback, Gradient Compression, and Why Adam Breaks It
Analyzes why gradient compression can cause Adam optimizer divergence in LLM training. · for ML engineers
devtoSteadyai76d ago - 59
The Loss Does Not See the Basis, but Adam Does
Loss basis issue · for ml engineers
arxivSteadyai92d ago