NVIDIA researcher confirms work on post-training chat models without breaking strengths
ziqiao_ma · x · 2026-11-08
Asked by Soumith Chintala and Giacomo how to post-train chat models without breaking their existing strengths, NVIDIA researcher Ziqiao Ma replied briefly: "We are working on it." Suggests related post-training work may surface soon.
More from Models
- JEPA seen as 'the big thing' in vision, but insiders say nobody actually cares — gabriberton · 2026-11-08
- Grok 4.7 scores 35, near-bottom on Artificial Analysis as critic slams Grok Bot pricing — Hesamation · 2026-11-08
- Claude Opus 4.5 Hits 70.2% on SWE-bench While Wiping Context After Every Command — cephaloform · 2026-11-08
- Grok on X web starts showing weekly usage limits as free traffic gets reined in — huangyun_122 · 2026-11-08
- Viral "finest LLMs" tier list ranks fictional future models from Gemini 4 Carbon to GPT-6 — iamfakhrealam · 2026-11-08
- Alibaba's Qoder Launches Stealth Model Qoduck, Free for Individual Users — Aiden_Tech_Ai · 2026-11-08