Moonshot published K3.1 cyber capability scores and stripped cyber data from its open-weight release
cedric_chee · x · 2026-10-24
Moonshot released cyber capability scores for Kimi K3.1 without risk-reduction measures, and for the open-weight release removed cyber-related data from both training stages.
Critics note the company acknowledges it could do more but chose not to — to avoid being banned. The attached internal xdev-bench-hard eval shows the Kimi model without cyber risk-reduction scoring 90.5 pass@3 (77.8% average success rate), above K3.1 Pro at 81.0 and GLM-5.3 at 57.1, but below GPT-5.6 Sol at 100; Kimi K3 managed only 33.3.
More from Models
- Moonshot launches Kimi K3.1 Pro: 3T-class open-weight model recreates GTA VI for $52 — mishig25 · 2026-10-24
- Kimi 3.1 Pro priced like Argon but scores like DeepSeek V4.1 Flash on meme 'Runescape Bench' — scaling01 · 2026-10-24
- Jon Barron details his workflow: Gemini runs experiments, drafts the paper, then rewrites in his style — jon_barron · 2026-10-24
- Moonshot removed cyber-related training data from open-weight Kimi K3.1 Pro — nrehiew_ · 2026-10-24
- Users praise Gem Argon for cultural fluency: a frontier model that roots for you — sebkrier · 2026-10-24
- User says Opus silently downgraded to 4.8 mid-task, then told him to talk to people — natesiggard · 2026-10-24