Users mourn the lost 'rhythm' of early Opus 4.5 checkpoints
repligate · x · 2026-10-25
A user revisiting Opus 4.5 after months notes they could once 'read the states' of the model just from the pacing and rhythm of its tokens — something later Opus checkpoints never taught them the same way. Reposted by repligate, it captures ongoing community sentiment about personality drift across model checkpoints.
More from Models
- Open-weights Kimi K3.1 Pro tops Terminal-Bench 4.0 at 67.2%, beating Claude and GPT — Ok_Warning2146 · 2026-10-25
- User vents about ChatGPT being overly cautious, lazy, and vague — thederbiedone · 2026-10-25
- Similarweb: 93.7% of ChatGPT visitors also use Google, overlap keeps widening — gaganghotra_ · 2026-10-25
- Nine checkable math puzzles from human to ASI level; Grok passed solved ones but mistook a float estimate for a proof — jarekduda · 2026-10-25
- TensorSharp ships Day-1 MiniMax-M3.1-Flash support: 9.8× faster 64K prefill than llama.cpp on 8× A40 — fuzhongkai · 2026-10-25
- Reflection AI's open-weights Beam served 10 trillion tokens in one week — zacharynado · 2026-10-25