Moonshot removed cyber-related training data from open-weight Kimi K3.1 Pro
nrehiew_ · x · 2026-10-24
Moonshot disclosed new cybersecurity controls for Kimi K3.1 Pro: as post-training scaled, vulnerability research capability steadily improved, traced to cyber content embedded in the general mid-training corpus and RL environment mix.
Internal experiments showed downstream cyber capability would rise substantially without data removal at either stage. For the open-weight release, the team identified and removed cyber-related data from both training stages while keeping general-purpose coding tasks, and says it will continue applying such controls at every training stage.
More from Models
- Terminal-Bench exposes NVIDIA Nemotron 3.5 near-zero score vs Kimi K3.1's 65.7% — JJitsev · 2026-10-24
- Bench shows GPT-5.6/6 Luna crush multi-agent coordination while DeepSeek V4 fails — JoshPurtell · 2026-10-24
- Grok Bot adds secure vault so agents can log in and pay without seeing your secrets — XFreeze · 2026-10-24
- Kimi K3.1 Pro reportedly first Chinese model to solve 20x20 nonograms — fairydreaming · 2026-10-24
- Kimi K3.1 vs GPT-6 Astra head-to-head on an identical agentic 3D task — GapNew4766 · 2026-10-24
- Kimi K3.1 Pro Launches: Scores 55 on Artificial Analysis, Open Weights Promised by Nov 2026 — Simon Willison · 2026-10-24