It has been hard to keep up with the flurry of strong agentic open-source models coming out of Chinese labs recently, including Moonshot’s Kimi K2, Z.ai’s GLM 4.5, and Qwen3-Coder¹.

The schedule is up for the 2025 edition of the PyTorch conference, which is now at the Moscone West in San Francisco.

The Second Half – Shunyu Yao – 姚顺雨

Dao-AILab/quack: A Quirky Assortment of CuTe Kernels

Back in 2021, OpenAI nixed its robotics team, leading to comments on Hacker News like “Reinforcement learning itself is a dead-end on a road to AI”. Now, in 2025 we are surrounded by RL post-trained reasoning models and Mary Meeker is using the word “unprecedented” a lot. This kind of skepticism/hype overlap is very common right now, as Helen Toner breaks down in her excellent recent post/talk on unresolved questions in AI:

Statecraft recently re-ran an interview from 2023 with Jason Matheny, formerly of IARPA: https://www.statecraft.pub/p/how-to-predict-the-future-278

In May Nvidia shipped CuTe‑DSL, the Python library they teased at GTC earlier in the year that mirrors CUTLASS’s C++ tensor‑layout . Then, at the start of June, the ‑dev label disappeared (so presumably its production ready now). The pitch is simple: Write speed‑of‑light kernels from the comfort of Python.

What content should you include in an LLM prompt? Many interesting use cases (enterprise tools, coding assistants) have more content than they can handle at once, so you chunk it up, turn each chunk into a vector with some sentence‑encoder, and store those vectors in a database. Later you vector‑search, pull back the relevant chunks and feed them to the LLM — better known as the RAG pattern.

What do you need to do to get better performance and GPU efficiency out of your model? The GPU-oriented folks at Stanford recently published an early preview of the work they have been doing on the LLM generation of kernels: Surprisingly Fast AI-Generated Kernels We Didn’t Mean to Publish (Yet) – and they have a list:

Traditionally, GPU collective network operations were issued from the framework on a separate CUDA stream than the local computation kernel launches. This allowed overlapping comms and hiding most or all of the network latency. NCCL exposes collectives as fully implemented kernels, and there have been various derivitives such as AMD’s RCCL or Berkeley’s new UCCLproject, which is aiming to be a drop-in replacement better suited for large-scale GPU workloads¹. Earlier versions of this sent the networking via the host, and later developed towards GPU-to-GPU peer to peer over connections like NVlink, and direct communication between GPU and the NIC for scale out. But the actual coordination was driven by launches from the CPU.