Tag: LLM
IBM proves a quantum edge that today’s AI models cannot match
A 60-page paper shows shallow quantum circuits winning on one functional and one sampling task, though the hardware…
Alibaba previews Qwen4 architecture in a 6B-active open model
An experimental open-weight release sketches the design direction for Alibaba's next flagship.
Canonical funds PhD study on AI rewriting C codebases in Rust
Canonical is funding a Bristol PhD project on using LLMs to translate large C codebases into safe Rust.
Anonymous Ox Alpha model has the internet guessing its maker
A free mystery model on OpenRouter has the internet guessing who built it.
Kog bets software unlocks 30x faster inference on stock GPUs
The French startup says its software can push 30x faster decoding out of the GPUs enterprises already own.
DeepSeek graduates V4 Pro to production after a four-month preview
DeepSeek's flagship V4 Pro reached general availability as build 0813, ending a four-month preview with unchanged pricing.
OpenAI swaps ChatGPT free tier to Luna with a new think button
OpenAI is dropping text chat caps and making GPT-5.6 Luna the default for free users.
Meta’s new Muse Code agent plans, codes, and checks its own work
Meta's first coding agent pairs with the co-trained Muse Spark 1.2 model in a beta launch.
Tencent’s 295B-parameter Hy3 heads overseas with Apache 2.0 weights
Hy3 lands on WorkBuddy, Miora, and Tencent Cloud as an open-weight option for developers.
AMD’s open 16B MoE model tops fully open rivals on benchmarks
AMD's fully open mixture-of-experts model leads open rivals while activating just 2.8B parameters.
AMD opens Instella MoE model trained on its Instinct GPUs
AMD ships Instella-MoE-16B-A3B, an open mixture-of-experts model trained on Instinct GPUs with 2.8B active parameters.
Chrome patched 1,072 flaws in June as AI hunting accelerates
Google fixed more Chrome flaws in June than in the prior two years, thanks to AI tools.