pretraining technique

Daily cross-source mention tracking for pretraining: how many AI news articles, Hacker News stories, research papers and GitHub trending repos mention it. Deterministic whole-word counts of real stored items, updated daily.

2 mentions, last 7 days 17 mentions, all stored days tracked since 2026-06-19
Daily total mentions · 2026-06-19 to 2026-07-09
0232026-06-19: 2 mentions2026-06-20: 1 mention2026-06-21: 1 mention2026-06-22: 1 mention2026-06-23: 3 mentions2026-06-24: 2 mentions2026-06-30: 1 mention2026-07-01: 2 mentions2026-07-02: 1 mention2026-07-03: 1 mention2026-07-07: 1 mention2026-07-09: 1 mention06-1906-3007-09
Per-source breakdown
SourceLast 7 daysAll stored days
News articles 0 0
Hacker News stories 0 0
Research papers 2 17
GitHub trending repos 0 0
Receipts · current items mentioning pretraining
paper Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence 2026-07-09 · external ↗ paper Vision Pretraining for Dense Spatial Perception 2026-07-07 · external ↗
How this is counted

Each item counts once per day if its title (plus summary or description where available) contains pretraining or one of its tracked aliases, matched case-insensitively on whole words. Heat on the Hype Meter board is scored as: raw = total mentions across sources in trailing 7 days; breadth = number of distinct sources with >=1 mention in those 7 days (1-4); heat = percentile rank of (raw * breadth) among all terms with raw>0, scaled 0-100, rounded. News and repo counts start the day tracking began; Hacker News and paper counts are backfilled from stored daily snapshots. No language model touches these numbers.

← Full hype board Model usage trends → AI news →