録 实録译稿

vllm-torchtpu is a new vLLM backend for Google Cloud TPUs built on TorchTPU, with Pallas kernels, multi-host parallelism, KV cache offloading, and disaggregated serving.

The geopolitics of open-weight AI are accelerating again this fall, as we would expect with the risks of all AI systems (open and closed) rising rapidly. We’ve seen many news items in the last week showing both the U.S. and Chinese governments positioning themselves for…

In the summer of 2010, postdoc Chris White worked with a multidisciplinary team at the Defense Advanced Research Projects Agency, or DARPA, to determine if a new approach to data analysis could prevent violence and protect troops during the war in Afghanistan. At the conclusion…

main.content table { display:block; width:100%; max-width:100%; overflow-x:auto; } #evaluation-prompt pre { background:var(--bs-body-bg); color:var(--bs-body-color); padding:18px; border:1px solid var(--bs-border-color); border-radius:6px; } #evaluation-prompt pre code {…

Brain encoding models aim to predict neural responses to stimuli. The most widely used approach fits readouts from frozen pretrained backbones. Using neural data to train large models from scratch or fully fine-tune them after task-pretraining has proven difficult because neural…

Reconstructing human activity from egocentric video requires more than recovering individual 3D components: the observer's motion, surrounding geometry, and hand-object interactions must fit together in a consistent world. Achieving this coherence requires both joint…

Human egocentric experience offers a rich foundation for learning how the world is organized and how it changes through interaction. Realizing this potential at scale requires turning growing volumes of recordings into learning resources that connect scene structure, object…

AI coding tools can generate code quickly, but speed alone does not solve the hardest software engineering problems. Developers still need to provide context, validate results, protect maintainability, and remain accountable for what ships. To learn how experienced developers…

We asked an internal research Claude model to design minibinders (small proteins that attach tightly to a target) against scorpion neurotoxins.

openbmb/tiny-processor-minicpm-ocr New model upload by openbmb on HuggingFace. Tags: minicpm_ocr, region:us.

Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Here’s a speech I gave on Saturday the 17th of October at the IMF meeting in Bangkok.…

Memory-native evolution requires making memory an intrinsic capability of foundation models, enabling them to continuously accumulate and utilize information. To this end, we introduce MemT-1.0, a new architecture of memory foundation models that internalize memory as part of…

Experience generated during deployment offers an opportunity for robots to improve policy reliability through learning from their own executions. However, learning through human corrections requires teleoperators to monitor execution and intervene when policies fail, a demand…

Runbooks already lagged behind code, and AI-assisted development is widening the gap. A former SRE explains how Code Search, Deep Search, and MCP ground incident response in the code itself.

How asynchronous preparation and NIXL EP enabled near-zero-downtime Elastic EP scaling in vLLM.

My obituary for programming Jeremy Scheff poignantly captures the key loss caused by coding agents for career programmers who enjoyed the puzzle of developing code: Programming turned into one of the greatest jobs you could imagine. You got paid to work on tricky little logic…

Learn how OpenAI is improving ChatGPT across low-resource programming languages, with new evaluations for software engineering, SQLite, and COBOL migration.

Today, we're announcing a preview of a new class of System One-style models: extraction models. Our first one, Clairic, is available hosted on Workers AI as @cf/cloudflare/clairic. We intend to release the model open-weights soon. Clairic is an extraction model that is able…

We benchmarked NVIDIA's new Arm-based Vera CPU on agentic Sandbox workloads. Under heavy load, it completes far more tasks than the x86 and Graviton alternatives, without sacrificing latency.

Post-train open LLMs like Qwen, GLM and Kimi into custom models you own, with SFT or RL. You write the training loop. Together runs the hard infrastructure.