Models & Labs
OpenAI addresses goblin references in models
The Verge AImedium confidence
Why it matters
- →Understanding these quirks in AI behavior is crucial for developers and researchers working with language models.
OpenAI has acknowledged a peculiar trend in its models, particularly the GPT-5.1, where they avoid discussing goblins and similar creatures. This issue was highlighted after a Wired report and OpenAI's subsequent explanation on their website.
Read originalMore from The Verge AI
© The Verge AIMarket & Regulationbusiness
Amazon Research Spurs White House Ban on Anthropic Fable
Amazon's cybersecurity research has reportedly driven the White House to restrict access to Anthropic's Fable 5 and Mythos 5 models. The findings suggested these models could be manipulated to provide information useful for cyberattacks, leading to discussions between Amazon CEO Andy Jassy and government officials. This decision has significant implications, as it prevents foreign nationals, including some of Anthropic's own researchers, from accessing these AI models. The situation underscores ongoing tensions between Anthropic and the U.S. government, particularly regarding the use of AI for surveillance and military purposes.
The Verge AI
© The Verge AIMarket & Regulationmodels
Anthropic halts Fable 5, Mythos 5 access by order
Anthropic has been forced to restrict access to its Fable 5 and Mythos 5 models due to a government mandate over national security issues. This restriction applies to all foreign nationals, impacting users globally, including Anthropic's own employees. The company expressed dissatisfaction with the government's lack of detailed information on the alleged security threats, which were communicated verbally. In response, Anthropic has collaborated with US and UK authorities to bolster security and revised its data retention policies. This development underscores the complex relationship between AI advancements and regulatory frameworks, as companies navigate compliance while maintaining innovation.
The Verge AI
© The Verge AIVideo & Creative AIimage
Apple's AI Photo Editing Tools in iOS 27
Apple's iOS 27 introduces new AI photo editing features that mark a significant shift for iPhone users. The updated Clean Up tool now effectively removes unwanted elements from photos by leveraging cloud-based AI models, aligning with Google's long-standing capabilities. Extend allows users to expand photo edges, adding plausible filler, while Spatial Reframing offers a 3D-like re-composition of images. These tools, though not as advanced as some competitors, represent a notable enhancement in Apple's native photo editing capabilities, potentially altering how users perceive and trust digital images.
The Verge AI
More in Models & Labs
Models & Labsmodels
vLLM v0.23.0 Release Enhances Model Support
The vLLM v0.23.0 release marks a significant step forward with enhancements across various components. DeepSeek-V4 has been optimized further, decoupling its metadata from previous versions and adding new attention kernels. Model Runner V2 now supports more dense models by default, improving performance for Llama and Mistral. The Rust frontend has matured with new endpoints and tool parsers, while compatibility with Transformers v5 ensures broader model support. These updates collectively enhance the robustness and versatility of vLLM, making it a more powerful tool for developers working with large language models.
vLLM Releases
Models & Labsmodels
Llama.cpp b9626 Release Adds Cohere2-MoE Support
The latest b9626 release of llama.cpp introduces architectural support for the cohere2-MoE model, marking a significant update for developers working with this model. This release also includes various technical improvements such as the removal of redundant checks and enhancements in tensor handling, which streamline the model's performance. By adding cohere2moe to the Llama Model Saver supported list, the update broadens the toolkit available for AI practitioners. While these changes may seem incremental, they collectively enhance the robustness and flexibility of llama.cpp, making it a more versatile tool for AI development.
llama.cpp Releases
Models & Labsmodels
llama.cpp b9627 Release Expands Platform Support
The b9627 release of llama.cpp continues to enhance its platform reach, though it doesn't introduce any groundbreaking features. This update includes support for a wide array of systems, from macOS and iOS to various Linux distributions and Windows configurations, including CUDA and Vulkan support. Notably, the release maintains its focus on making llama.cpp a versatile tool across different hardware setups, but it doesn't introduce new model architectures or quantization methods. This iteration is more about solidifying its presence across multiple operating systems rather than introducing novel capabilities.
llama.cpp Releases