A new release of the llama.cpp library, version b11773, includes support for the H2O-Lightning-4B model, which is a Qwen3.5-4B model with a merged decision LoRA. This update also features improvements to code redundancy and label management. Separately, a security audit engine called Chitos has been developed to automate AI security assessments, particularly for financial institutions, by mimicking attacker strategies and using AI for verification. Additionally, a dashboard named VERA has been created to help non-technical staff assess responsible AI practices in compliance with the EU AI Act, reducing assessment time and improving usefulness. AI
IMPACT New tools are emerging for AI security auditing and compliance, potentially streamlining assessments and improving responsible AI practices.
RANK_REASON The cluster includes updates to an open-source library (llama.cpp), a new AI security audit engine (Chitos), and a dashboard for assessing AI compliance (VERA).
Read on llama.cpp — Releases →
AI-generated summary · Google Gemini · from 9 sources. How we write summaries →
