Dev Observes astra: Terrible Judgement but Distinct Big-Model Vibes
louisvarge · x · 2026-10-05
Developer louisvarge shares a hands-on impression of the model astra: it's a really weird one — terrible judgement, yet it still carries a distinct "big model smell," suggesting a large parameter count with poor decision-making tuning. An informal first-hand behavioral observation from the AI community.
Related event: Developers find Astra's judgment weak, on par with Sol on complex tasks(3 posts)→
More from Models
- Open-weights security model apex-flash-1 lands $1M in bug bounties, tops HackerOne board — glenbeer · 2026-10-05
- Cursor CEO:前沿模型在复杂任务上集体表现相似且挣扎 — zeeg · 2026-10-05
- Gemini 4 Argon hands-on video sparks debate: is Google back? — RelevantEmergency707 · 2026-10-05
- GLM 5.3 flash on dual DGX Sparks gets 50-90% decode boost with new open recipe — swiebertjee · 2026-10-05
- Open-Weight Model Rumor Watchlist: Qwen 4 by November, Kimi K3.1 Eyed for October — kimmonismus · 2026-10-05
- GPT-5.6 Sol High keeps flip-flopping on answers, sycophancy complaints persist — PressPlayPlease7 · 2026-10-05