DeepSeek V4.1 Flash suspected silent update shows high code-edit failure rate
teortaxesTex · x · 2026-10-13
Users suspect DeepSeek V4.1 Flash received a behind-the-scenes update, and one reports the model now has an "atrocious" failure rate applying code edits — often missing or deleting chunks of code. Screenshots show a patch script failing midway, a lost declaration in textures.js, accidentally removed computation code needing re-insertion, and a shader compile error, with one session at 186 steps and 65.7M tokens. Cause unclear.
More from Models
- Scale's Alexandr Wang says muse-wonder bug fixes are already showing clear improvement — alexandr_wang · 2026-10-13
- Fine-tuning speech models for Algerian Darija: a team's four open questions — Fair_Bed_914 · 2026-10-13
- Google owns every dimension of personal context yet ships no unifying AI product — signulll · 2026-10-13
- Jason: 99% of enterprises don't need esoteric math prizes, open source models will win — markjeffrey · 2026-10-13
- Qwen3.8 Flash Next IQ3_S hits 90 tk/s locally, low-poly 3D generation beats expectations — migsperez · 2026-10-13
- Meta ships muse-wonder-full-1.0 to UK same-day as US, researcher says thanks — SamuelAlbanie · 2026-10-13