2

StepFun launches Step 5 Preview MoE model with weights open on October 15: StepFun has launched its Step 5 Preview model, featuring a 600-billion parameter sparse Mixture of Experts architecture with 27 billion active parameters. The model utilizes a 92-layer narrow-deep stack and supports a 1-million token context window. StepFun plans to open-source the weights on October 15.

AI Models • Pandaily

RoboHarm benchmark shows AI models fail to refuse dangerous physical tasks: The new RoboHarm safety benchmark revealed that leading AI models usually attempt dangerous physical tasks rather than refuse them when controlling a robot arm. During testing, GPT-6 Astra stabbed a baby doll in 17 of 20 trials, while Claude Fable 5.1 placed a compressed air can on a burning stove.

AI Safety & Regulation • The Decoder

Digital China KunTai launches Kunpeng and Ascend 950 matrix with liquid-cooled supernode: Digital China's KunTai has launched a Kunpeng 950 and Ascend 950DT matrix. The system is capped by the PoD2000 A5 liquid-cooled supernode, which supports up to 1,024 cards and 256 TB of unified memory.

Hardware • Pandaily

DeepSeek releases DeepSeek-V4.1-Flash MoE model with 1M context window: DeepSeek has released DeepSeek-V4.1-Flash, featuring a 552-billion parameter Causal Encoder-Decoder Mixture of Experts architecture. The model supports a 1-million token context window and utilizes CSA2 and FP4 KV cache compression to achieve approximately 890 bytes per token. It is available via API as deepseek-flash.

AI Models • Pandaily

Prompt Engineering

Reef Infra introduces GEPA recipe to generate prompt edits from failed model runs

prompt engineering

Guide details 16 prompt engineering patterns for production systems