Grok 4.7 delayed a few days as Musk blames RL length penalty for model giving up on hard tasks
minchoi · x · 2026-09-12
Elon Musk says Grok 4.7 needs a few more days, citing a likely RL training bug: the team may have penalized response length too much, so the model gives up too early on hard tasks it can actually solve and isn't rigorous enough checking its own work. A rare public admission tying model behavior defects to reward design tradeoffs.
Related event: Musk Confirms Grok 4.7 Delay, Blames Overweight RL Penalty on Reply Length(7 posts)→
More from Models
- Gemini 4 Argon Hits 1.73% WER on OCR Benchmark, Surpassing Human Experts — gabriberton · 2026-10-28
- Meta's Omnilingual MT covers 1,600 languages, 1B–8B models rival 70B LLMs — kalyan_kpl · 2026-10-28
- Uncensored LFM3-1.7B GGUF weights released on Hugging Face — kalyan_kpl · 2026-10-28
- Token demand on OpenRouter grew 27x to 165T weekly, open models now serve 72% — FinanceYF5 · 2026-10-28
- Tesla Optimus 'robot university' leak: 40-camera mocap suits, 24/7 data ops for months — Scobleizer · 2026-10-28
- Letta's FrontierMemory leaderboard: Letta Dream tops at 73.3%, beating OpenAI and Claude agents — afurgs · 2026-10-28