Model recipe via rigorous experiments, but scaling could rewrite the story
liliyu_lili · x · 2026-11-09
In a reply exchange with giffmana, the author notes the team landed on the best recipe through rigorous experiments and built a good model, but stresses that scaling and new research can change the story, so they keep an open mind about the future — framing scientific mindset and challenging old assumptions as the durable approach.
More from Models
- Running Ripple 65B Preview NVFP4 on SGLang with 2x RTX PRO 6000: full recipe — HankYeomans · 2026-11-09
- DSpine boosts Qwen3.8-27B to 94-132 tok/s in llama.cpp, 1.75x over DFlash2 — pbaylies · 2026-11-09
- Reflection launches Beam: 501B-param agentic open model trained with AdamC optimizer — andersonbcdefg · 2026-11-09
- Models can be trained to talk to each other in inscrutable neuralese text — Sauers_ · 2026-11-09
- Artificial Analysis launches Decision Index v2 with 45% private data and calibration-based scoring — ArtificialAnlys · 2026-11-09
- GLM-4.5 Hit #1 on Hugging Face Within 48 Hours of MIT-Licensed Release — ccerrato147 · 2026-11-09