Search “Meta”

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

AI startup Vectorial raises $3 million to use AI to simulate human behavior for consumer research

AI startup Vectorial has closed a $3 million seed round co-led by Smilegate and SenseAI Ventures to launch its consumer research platform that simulates human behavior. The platform is based on the behavioral model SAPIENS, whose training data covers more than 2.5 million real consumers; it predicts consumer behavior 30% more accurately than general-purpose models and achieves about 86% accuracy compared with real user research. Customers including GoodRx, Boat, Everkind and Concrete Software run an average of 600 simulations per month, and the company will use the new funding to build SAPIENS 2.0.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

PyTorch 2.15 Brings FP8 and Faster MoE Decoding to MPS

PyTorch 2.15 adds FP8 support to the MPS backend and makes MoE model decoding 5–6.5 times faster than 2.14: on an M3 Ultra with torch.compile, Qwen3.6-35B-A3B goes from 9.4 tokens/s to 61.3 tokens/s, and LFM2.5-8B-A1B goes from 23.6 to 123.6 tokens/s. For Wan2.2's tiled VAE video decoding, peak driver memory drops from about 270 GiB to 8 GiB, and the image corruption issue with Qwen-Image-2.1's VAE encoding on MPS is also fixed with a padding change, raising the encode-decode round-trip PSNR from 10.06 dB to 29.95 dB. This version requires macOS 15 or later and Python 3.11 or later.

English translation is being prepared.

English translation is being prepared.

Tolan raises $30 million in funding, AI companion app surpasses 250,000 users

AI companion app Tolan has completed a new $30 million funding round, with investors including Union Square Ventures, Khosla Ventures, and Homebrew, doubling its cumulative funding. PitchBook estimates its valuation at over $268 million. Tolan uses cartoon characters to help users manage daily tasks and handle relationships and emotions. It charges through a tiered membership model, and its annual recurring revenue has reached the "tens of millions of dollars" level. Over the past year, its active users have exceeded 250,000, with cumulative conversations surpassing 70 million. About 70% of users are women, and 40% are parents.

Ant Group's inclusionAI publishes the sglang_ling_v3 repository on GitHub

Ant Group's inclusionAI launches the sglang_ling_v3 repository on GitHub. Its contents are based on the open-source inference framework SGLang, which targets large language models, vision-language models, and diffusion models, and is optimized for agent workloads, large-scale serving, and RL rollout. The repository README shows that SGLang supports GPU, TPU, NPU, CPU, and Apple Silicon, and includes the built-in image and video generation engine SGLang Diffusion.

Ant Group's inclusionAI open-sources the Ming-3.0-audio-flash native audio model

Ant Group's inclusionAI has open-sourced the Ming-3.0-audio-flash native audio model on GitHub. It uses a 126B-parameter MoE architecture with 6.5B activated parameters, and there is also a Ming-3.0-audio-tiny version with 9.3B parameters and 2.7B activated parameters. The models cover Chinese and English recognition across multiple dialects and accents, long audio understanding, audio reasoning, and end-to-end speech dialogue. They support three inference methods: vLLM-Omni, SGLang, and Transformers, and are released under the MIT license.

Former engineer accused of planting a listening device in an OpenAI executive conference room is ordered to remain in custody over Saudi ties

A U.S. federal judge ruled that Abdulrahman Alfozan, a former engineer accused of hiding a recording device in an OpenAI executive conference room, will remain in custody before trial, citing his ties to Saudi Arabia as creating a flight risk. Prosecutors said the Saudi government was willing to fund his bail unconditionally, and alleged that between 2025 and 2026 he communicated frequently with a member of the Saudi royal family and was recruited to return home to take part in an AI initiative. Alfozan had worked at OpenAI for just over a year; the device was discovered in late September, and he was arrested by the FBI this month after being fired. He currently faces one count of intercepting communications, which carries a maximum sentence of 5 years in prison.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

One Search, Two Layers of Optimization: PyTorch Helion Introduces Compiler Controls with NVIDIA CompileIQ

PyTorch Helion can now search ptxas compiler control parameters through NVIDIA CompileIQ, jointly with the existing kernel configuration search, further tapping performance headroom on the roughly 27 kernels that have already been pre-tuned. CompileIQ packages tuning results as a portable Advanced Control File (ACF); Helion supports specifying a fixed ACF, and also supports searching multiple ACFs together as one dimension of autotuning. These compiler controls are exposed by ptxas starting from CUDA 13.3.

Illinois Governor JB Pritzker on AI: It could help people live to 100, or threaten humanity like nuclear weapons

In an appearance on the Decoder podcast, Illinois Governor JB Pritzker said AI could both help humans live to 100 and pose a threat like nuclear weapons. He believes data centers, privacy, and antitrust are sparking bipartisan backlash, and that the government should do more to protect the public rather than just working with companies.