Data filtering emerges as its own scaling lever as researchers debate quality-scaling limits
tobyordoxford · x · 2026-09-09
In a reply thread, David Manheim argues marginal evidence suggests more model improvement comes from causes other than "add compute" or "add data," making those trajectories flatter than claimed — and raising the question of where quality-scaling tops out. Toby Ord agrees that breaking out data filtering as its own category is a useful mental-model update.
Related event: Scholars Debate Whether Pretraining Scaling Is Slowing Down(4 posts)→
More from Research
- DeepMind paper: LLMs can't sample randomly on their own, only convert given seeds — rohanpaul_ai · 2026-10-27
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- World Model Assigns Weight and Friction to Objects, Then Simulates with a Physics Engine — jonstephens85 · 2026-10-27
- Token Collapse: Study Finds Graph Soft Prompting Tokens in LLMs Fail to Encode Structure — Sapienza · 2026-10-27
- Mathathon 2.0 launches with Terence Tao's SAIR, $1.5M in compute for teams — _sathvikr · 2026-10-27
- The 'sweet lesson' is real: two kinds of ML problems need opposite strategies — mike64_t · 2026-10-27