{"id":9625,"date":"2026-05-21T19:23:00","date_gmt":"2026-05-21T17:23:00","guid":{"rendered":"http:\/\/stocks-future.com\/?guid=96bf77e31a150efd14b26f2ec615b04a"},"modified":"2026-05-21T19:23:00","modified_gmt":"2026-05-21T17:23:00","slug":"google-deepmind-features-hirundos-security-hardened-gemma-4-model-outperforms-llms-170x-its-size-on-security","status":"publish","type":"post","link":"https:\/\/stocks-future.com\/?p=9625","title":{"rendered":"Google DeepMind Features Hirundo\u2019s Security-Hardened Gemma 4 Model \u2013 Outperforms LLMs 170x Its Size on Security"},"content":{"rendered":"

TEL AVIV, Israel--(BUSINESS WIRE)--Google DeepMind<\/a> has featured<\/a> Hirundo\u2019s security-hardened variant of Gemma 4 in its Gemmaverse \u2013 the official showcase for the Gemma open-model ecosystem. The feature validates Hirundo<\/a>\u2019s weight-level machine unlearning approach as a production-grade solution to one of the most persistent vulnerabilities in enterprise AI deployments: prompt injection attacks.<\/p><\/a><\/a><\/a><\/a>

\nThe hardened model is built on Google\u2019s Gemma 4 E4B instruction-tuned base. At 4 billion parameters, it outperforms models more than 170 times its size on prompt injection resistance \u2013 including DeepSeek V3.2-Exp (685B), GPT-OSS-120B, and Qwen3-235B \u2013 while preserving full utility across standard benchmarks.<\/p>

\nAI security is a behavioral problem, not a size problem<\/b><\/p>

\nPrompt injection \u2013 where adversarial inputs manipulate a model into overriding its system instructions \u2013 remains one of the most damaging attack vectors in production Large Language Model (LLM) deployments. The prevailing assumption has been that larger models are inherently more robust. Hirundo\u2019s results directly contradict that.<\/p>

\nRather than applying external filters or inference-time guardrails, Hirundo\u2019s platform targets and removes the specific model weights responsible for susceptibility to adversarial manipulation. The model effectively forgets the behaviors that enable attacks \u2013 at the weight level, not the prompt level. Despite those behaviors usually being tied to instruction-following, a desired trait of LLMs, Hirundo\u2019s technology preserved general instruction following while mitigating prompt injection susceptibility.<\/p>

\nThe results against industry-leading open-weight models, benchmarked using Meta\u2019s PurpleLlama CyberSecEval dataset:<\/p>