https://24-ai.news/hr/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI hr 2026-07-06T00:00:00Z AMD Quark donosi MXFP4 kvantizaciju za FLUX.1-dev na Instinct MI350 uz 1,92× ubrzanje https://24-ai.news/hr/news/2026-07-06/amd-primus-tuning-agent/ 24 AI hr 2026-07-06T00:00:00Z AMD Primus Tuning Agent automatski pronalazi optimalnu LLM konfiguraciju za MI355X klastere https://24-ai.news/hr/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI hr 2026-07-06T00:00:00Z Alberta skenirala 466 milijuna linija koda za 20 sati: vlada koristi Claude za kibernetičku sigurnost https://24-ai.news/hr/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI hr 2026-07-06T00:00:00Z Anthropic otkrio J-space: emergentni unutarnji radni prostor unutar Claudea https://24-ai.news/hr/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI hr 2026-07-06T00:00:00Z MiniMax M2, M2.1 i M2.5 dostupni na Amazon Bedrocku https://24-ai.news/hr/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI hr 2026-07-06T00:00:00Z AWS automatizira redigiranje PII-ja u slikama s Nova 2 Lite i SAM 3 https://24-ai.news/hr/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI hr 2026-07-06T00:00:00Z AWS uvodi rDPO — selektivno zaboravljanje za Amazon Nova modele https://24-ai.news/hr/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI hr 2026-07-06T00:00:00Z AI tvornice postaju nacionalna infrastruktura: kako države grade suverenost u AI eri https://24-ai.news/hr/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI hr 2026-07-06T00:00:00Z NVIDIA-ini otvoreni modeli dominiraju ICML-om 2026 u Seulu https://24-ai.news/hr/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI hr 2026-07-06T00:00:00Z Kvantne tehnologije i AI: što je realno danas, a što ostaje budućnost https://24-ai.news/hr/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI hr 2026-07-06T00:00:00Z OpenAI objavio GPT-Realtime-2.1 i mini varijantu: poboljšano prepoznavanje glasa i rukovanje šumom https://24-ai.news/hr/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI hr 2026-07-06T00:00:00Z vLLM i Tencent Hunyuan upstreamaju dva HPC backenda za NVIDIA Hopper GPU-ove https://24-ai.news/en/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI en 2026-07-06T00:00:00Z AMD Quark delivers MXFP4 quantization for FLUX.1-dev on Instinct MI350 with 1.92x speedup https://24-ai.news/en/news/2026-07-06/amd-primus-tuning-agent/ 24 AI en 2026-07-06T00:00:00Z AMD Primus Tuning Agent automatically finds the optimal LLM configuration for MI355X clusters https://24-ai.news/en/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI en 2026-07-06T00:00:00Z Alberta scanned 466 million lines of code in 20 hours: government uses Claude for cybersecurity https://24-ai.news/en/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI en 2026-07-06T00:00:00Z Anthropic discovers J-space: an emergent internal workspace inside Claude https://24-ai.news/en/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI en 2026-07-06T00:00:00Z MiniMax M2, M2.1, and M2.5 now available on Amazon Bedrock https://24-ai.news/en/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI en 2026-07-06T00:00:00Z AWS automates PII redaction in images with Nova 2 Lite and SAM 3 https://24-ai.news/en/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI en 2026-07-06T00:00:00Z AWS introduces rDPO — selective unlearning for Amazon Nova models https://24-ai.news/en/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI en 2026-07-06T00:00:00Z AI factories become national infrastructure: how nations build sovereignty in the AI era https://24-ai.news/en/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI en 2026-07-06T00:00:00Z NVIDIA's open models dominate ICML 2026 in Seoul https://24-ai.news/en/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI en 2026-07-06T00:00:00Z Quantum technologies and AI: what is realistic today, and what remains the future https://24-ai.news/en/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI en 2026-07-06T00:00:00Z OpenAI releases GPT-Realtime-2.1 and mini variant: improved voice recognition and noise handling https://24-ai.news/en/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI en 2026-07-06T00:00:00Z vLLM and Tencent Hunyuan upstream two HPC backends for NVIDIA Hopper GPUs https://24-ai.news/de/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI de 2026-07-06T00:00:00Z AMD Quark bringt MXFP4-Quantisierung für FLUX.1-dev auf Instinct MI350 mit 1,92× Beschleunigung https://24-ai.news/de/news/2026-07-06/amd-primus-tuning-agent/ 24 AI de 2026-07-06T00:00:00Z AMD Primus Tuning Agent findet automatisch die optimale LLM-Konfiguration für MI355X-Cluster https://24-ai.news/de/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI de 2026-07-06T00:00:00Z Alberta scannte 466 Millionen Codezeilen in 20 Stunden: Regierung nutzt Claude für Cybersicherheit https://24-ai.news/de/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI de 2026-07-06T00:00:00Z Anthropic entdeckt J-space: emergenter interner Arbeitsraum in Claude https://24-ai.news/de/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI de 2026-07-06T00:00:00Z MiniMax M2, M2.1 und M2.5 jetzt auf Amazon Bedrock verfügbar https://24-ai.news/de/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI de 2026-07-06T00:00:00Z AWS automatisiert PII-Schwärzung in Bildern mit Nova 2 Lite und SAM 3 https://24-ai.news/de/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI de 2026-07-06T00:00:00Z AWS führt rDPO ein – selektives Vergessen für Amazon-Nova-Modelle https://24-ai.news/de/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI de 2026-07-06T00:00:00Z KI-Fabriken als nationale Infrastruktur: Wie Staaten Souveränität im KI-Zeitalter aufbauen https://24-ai.news/de/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI de 2026-07-06T00:00:00Z NVIDIAs offene Modelle dominieren die ICML 2026 in Seoul https://24-ai.news/de/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI de 2026-07-06T00:00:00Z Quantentechnologien und KI: Was heute realistisch ist und was Zukunft bleibt https://24-ai.news/de/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI de 2026-07-06T00:00:00Z OpenAI veröffentlicht GPT-Realtime-2.1 und Mini-Variante: verbesserte Spracherkennung und Rauschunterdrückung https://24-ai.news/de/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI de 2026-07-06T00:00:00Z vLLM und Tencent Hunyuan streamen zwei HPC-Backends für NVIDIA-Hopper-GPUs upstream https://24-ai.news/zh/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI zh 2026-07-06T00:00:00Z AMD Quark在Instinct MI350上以MXFP4量化FLUX.1-dev，实现1.92倍加速 https://24-ai.news/zh/news/2026-07-06/amd-primus-tuning-agent/ 24 AI zh 2026-07-06T00:00:00Z AMD Primus Tuning Agent自动为MI355X集群寻找最优LLM配置 https://24-ai.news/zh/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI zh 2026-07-06T00:00:00Z 艾伯塔省20小时扫描4.66亿行代码：政府使用Claude保障网络安全 https://24-ai.news/zh/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI zh 2026-07-06T00:00:00Z Anthropic揭示J-space：Claude内部涌现的全局工作空间 https://24-ai.news/zh/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI zh 2026-07-06T00:00:00Z MiniMax M2、M2.1和M2.5登陆Amazon Bedrock https://24-ai.news/zh/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI zh 2026-07-06T00:00:00Z AWS使用Nova 2 Lite和SAM 3自动编辑图像中的个人信息 https://24-ai.news/zh/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI zh 2026-07-06T00:00:00Z AWS引入rDPO——为Amazon Nova模型提供选择性遗忘功能 https://24-ai.news/zh/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI zh 2026-07-06T00:00:00Z AI工厂成为国家基础设施：各国如何在AI时代建立主权 https://24-ai.news/zh/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI zh 2026-07-06T00:00:00Z NVIDIA开放模型主导2026年首尔ICML大会 https://24-ai.news/zh/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI zh 2026-07-06T00:00:00Z 量子技术与AI：当下已成现实与仍属未来的边界 https://24-ai.news/zh/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI zh 2026-07-06T00:00:00Z OpenAI发布GPT-Realtime-2.1及mini版本：改进语音识别与噪音处理 https://24-ai.news/zh/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI zh 2026-07-06T00:00:00Z vLLM与腾讯混元联合上游两个HPC后端，支持NVIDIA Hopper GPU https://24-ai.news/ja/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI ja 2026-07-06T00:00:00Z AMD QuarkがInstinct MI350向けFLUX.1-devのMXFP4量子化で1.92倍高速化を実現 https://24-ai.news/ja/news/2026-07-06/amd-primus-tuning-agent/ 24 AI ja 2026-07-06T00:00:00Z AMD Primus Tuning AgentがMI355Xクラスター向けの最適LLM設定を自動発見 https://24-ai.news/ja/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI ja 2026-07-06T00:00:00Z 4億6600万行のコードを20時間でスキャン：アルバータ州政府がClaudeでサイバーセキュリティを強化 https://24-ai.news/ja/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI ja 2026-07-06T00:00:00Z AnthropicがJ-spaceを発見：Claude内部に出現した創発的「グローバル・ワークスペース」 https://24-ai.news/ja/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI ja 2026-07-06T00:00:00Z MiniMax M2・M2.1・M2.5がAmazon Bedrockで利用可能に https://24-ai.news/ja/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI ja 2026-07-06T00:00:00Z AWSがNova 2 LiteとSAM 3で画像内の個人情報を自動墨消し https://24-ai.news/ja/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI ja 2026-07-06T00:00:00Z AWSがAmazon Novaモデルに選択的アンラーニング手法rDPOを導入 https://24-ai.news/ja/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI ja 2026-07-06T00:00:00Z AI工場が国家インフラとなる：AI時代の主権をめぐる各国の取り組み https://24-ai.news/ja/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI ja 2026-07-06T00:00:00Z NVIDIAのオープンモデルがソウルのICML 2026を席巻 https://24-ai.news/ja/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI ja 2026-07-06T00:00:00Z 量子技術とAI：今日現実的なものと未来にとどまるもの https://24-ai.news/ja/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI ja 2026-07-06T00:00:00Z OpenAIがGPT-Realtime-2.1とminiバリアントを発表：音声認識と雑音処理を改善 https://24-ai.news/ja/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI ja 2026-07-06T00:00:00Z vLLMとTencent HunyuanがNVIDIA Hopper GPU向け2つのHPCバックエンドをアップストリーム https://24-ai.news/ko/news/2026-07-06/amd-mxfp4-quark-flux-diffusers/ 24 AI ko 2026-07-06T00:00:00Z AMD Quark, Instinct MI350에서 FLUX.1-dev를 MXFP4로 양자화해 1.92배 속도 향상 https://24-ai.news/ko/news/2026-07-06/amd-primus-tuning-agent/ 24 AI ko 2026-07-06T00:00:00Z AMD Primus Tuning Agent, MI355X 클러스터에서 최적의 LLM 구성을 자동으로 탐색 https://24-ai.news/ko/news/2026-07-06/anthropic-alberta-government-cybersecurity/ 24 AI ko 2026-07-06T00:00:00Z 앨버타, 20시간 만에 4억 6,600만 줄 코드 스캔: 정부의 사이버 보안에 Claude 활용 https://24-ai.news/ko/news/2026-07-06/anthropic-global-workspace-jspace/ 24 AI ko 2026-07-06T00:00:00Z Anthropic, Claude 내부에서 긴급 출현한 J-space 공개: 모델의 내부 작업 공간 https://24-ai.news/ko/news/2026-07-06/aws-minimax-models-bedrock/ 24 AI ko 2026-07-06T00:00:00Z MiniMax M2, M2.1, M2.5, Amazon Bedrock에서 사용 가능 https://24-ai.news/ko/news/2026-07-06/aws-nova-pii-image-redaction/ 24 AI ko 2026-07-06T00:00:00Z AWS, Nova 2 Lite와 SAM 3으로 이미지 내 개인정보 자동 삭제 자동화 https://24-ai.news/ko/news/2026-07-06/aws-nova-selective-unlearning-rdpo/ 24 AI ko 2026-07-06T00:00:00Z AWS, Amazon Nova 모델에 선택적 언러닝 기술 rDPO 도입 https://24-ai.news/ko/news/2026-07-06/nvidia-nations-ai-strategic-priorities/ 24 AI ko 2026-07-06T00:00:00Z AI 공장이 국가 인프라로: 국가들이 AI 시대에 주권을 구축하는 방법 https://24-ai.news/ko/news/2026-07-06/nvidia-open-models-icml-2026/ 24 AI ko 2026-07-06T00:00:00Z NVIDIA 오픈 모델, 서울 ICML 2026 석권 https://24-ai.news/ko/news/2026-07-06/oecd-quantum-technologies-ai/ 24 AI ko 2026-07-06T00:00:00Z 양자 기술과 AI: 오늘날 현실과 미래의 가능성 https://24-ai.news/ko/news/2026-07-06/openai-gpt-realtime-2-1/ 24 AI ko 2026-07-06T00:00:00Z OpenAI, GPT-Realtime-2.1과 mini 모델 공개: 음성 인식과 노이즈 처리 개선 https://24-ai.news/ko/news/2026-07-06/vllm-hunyuan-hpc-ops-backends/ 24 AI ko 2026-07-06T00:00:00Z vLLM과 Tencent Hunyuan, NVIDIA Hopper GPU를 위한 두 가지 HPC 백엔드 업스트림 https://24-ai.news/hr/news/2026-07-04/apple-icml-2026-roundup/ 24 AI hr 2026-07-04T00:00:00Z Apple na ICML 2026: 28 radova, on-device MLX i negativan rezultat o multi-agentima https://24-ai.news/en/news/2026-07-04/apple-icml-2026-roundup/ 24 AI en 2026-07-04T00:00:00Z Apple at ICML 2026: 28 Papers, On-Device MLX and a Negative Result on Multi-Agents https://24-ai.news/de/news/2026-07-04/apple-icml-2026-roundup/ 24 AI de 2026-07-04T00:00:00Z Apple auf der ICML 2026: 28 Paper, On-Device-MLX und ein negatives Ergebnis zu Multi-Agenten https://24-ai.news/zh/news/2026-07-04/apple-icml-2026-roundup/ 24 AI zh 2026-07-04T00:00:00Z 苹果亮相ICML 2026：28篇论文、端侧MLX与多智能体负面结论 https://24-ai.news/ja/news/2026-07-04/apple-icml-2026-roundup/ 24 AI ja 2026-07-04T00:00:00Z AppleのICML 2026：28本の論文、オンデバイスMLX、マルチエージェントへの否定的知見 https://24-ai.news/ko/news/2026-07-04/apple-icml-2026-roundup/ 24 AI ko 2026-07-04T00:00:00Z Apple의 ICML 2026: 논문 28편, 온디바이스 MLX, 멀티 에이전트 부정적 결과