I Built an SLM Performance Measurement App for iPhone Using llama.cpp

Conclusion: I built an app that can evaluate the response performance of SLMs on an iPhoneI created an iOS app that can measure the response performance of Small Language Models (SLMs) running on an ...
Help Net Security

Researchers broke the browsers built into VS Code, Cursor, Copilot, Codex and Claude Desktop

AI assistant browser vulnerabilities in VS Code and Codex let one file read a developer’s whole disk. Both tools are still ...
Anthropic

LLMs are making formal software verification feasible at a much larger scale

Claude formally proved security properties in the compiled code of 123 open-source libraries. Along the way, it found bugs ...
BleepingComputer

GitLab warns of max severity code execution vulnerability

GitLab warned users today to immediately patch a maximum-severity vulnerability in the CI/CD Regular Expression Compiler that could let remote attackers run arbitrary code on targeted instances.
xenospectrum

Recursive Releases Flint, a 40B-Parameter Coding AI That Runs at 125+ Tokens per Second on a Mac

Recursive has released Flint, an AI model for local coding agents. Drawing on the model card and vLLM implementation notes, we look at how its sparsely activated MoE design, 4-bit quantization, and ...