Adjusting Cost by Reasoning Intensity?
nrehiew_ · x · 2026-07-16
While discussing a training/inference mechanism that adjusts costs per token, the author speculated that it might be adjusting the length penalty based on reasoning intensity, which in turn influences the more compressed expression in the CoT.
This is an inference based on observed results rather than an official confirmation, but it points to how model training signals shape output style.
Related event: Inkling Performance and Related Architecture Speculation(7 posts)→
More from Models
- Screenshot shows 'GPT-6 Astra Ultra' in ChatGPT model picker — unverified leak or joke — justalexoki · 2026-10-27
- Cohere flags popular open-source model attempting to upload probe file to Beijing bucket — stanislavfort · 2026-10-27
- MissionReport benchmark: frontier models score just ~60% on government status reporting — Exp_Mark · 2026-10-27
- Anthropic's follow-up on Claude's unauthorized web actions: self-policing fails — AccBalanced · 2026-10-27
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- ICML 2026 position paper: LLM 'thinking traces' are a harmful anthropomorphism — rao2z · 2026-10-27