https://24-ai.news/hr/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI hr 2026-06-20T00:00:00Z AMD: ROCm optimizacija Matrix3D-a za 3D svjetove ubrzala renderiranje do 54 posto na Instinct GPU-ima https://24-ai.news/hr/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI hr 2026-06-20T00:00:00Z Anthropic: Claude Code v2.1.183 blokira destruktivne git i infrastrukturne naredbe u auto modu https://24-ai.news/hr/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI hr 2026-06-20T00:00:00Z arXiv:2606.20205: psihološki profili jezičnih modela uglavnom su mjerni artefakt, a ne stabilna osobnost https://24-ai.news/hr/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI hr 2026-06-20T00:00:00Z arXiv:2606.20508: što jezični modeli uče iz miješanih demonstracija sigurnog i štetnog ponašanja https://24-ai.news/hr/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI hr 2026-06-20T00:00:00Z arXiv:2606.20517: Multi-LCB proširuje LiveCodeBench na 12 programskih jezika i otkriva Python overfitting kod 24 modela https://24-ai.news/hr/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI hr 2026-06-20T00:00:00Z arXiv:2606.19808: SEVRA selektivnom verifikacijom štedi do 91 posto tokena u rasuđivanju modela https://24-ai.news/hr/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI hr 2026-06-20T00:00:00Z arXiv:2606.20333: SoftSkill komprimira skill-dokumente u 32 latentna tokena i diže LiveMath za 42,1 boda https://24-ai.news/hr/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI hr 2026-06-20T00:00:00Z AWS: SageMaker dobiva preko 100 detaljnih inference metrika i Insights nadzornu ploču na CloudWatchu https://24-ai.news/hr/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI hr 2026-06-20T00:00:00Z EU AI Office: konzorcij EUROPA gradi otvoreni europski frontier model s preko 400 milijardi parametara za svih 24 jezika EU-a https://24-ai.news/hr/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI hr 2026-06-20T00:00:00Z GitHub: Copilot gasi Opus 4.6 (fast) 29. lipnja, dodaje AGENTS.md u code review i polje ai_credits_used u API https://24-ai.news/hr/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI hr 2026-06-20T00:00:00Z GitHub: interni analitički agent Qubot smanjio vrijeme rješavanja upita za oko 66 posto https://24-ai.news/hr/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI hr 2026-06-20T00:00:00Z NVIDIA: partneri na Cannes Lions 2026 pokazali agentski AI za marketing, Criteo udvostručio brzinu treninga na Blackwellu https://24-ai.news/hr/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI hr 2026-06-20T00:00:00Z UK AISI: Engineering Playbook otvara infrastrukturu za evaluaciju frontier modela u pet slojeva https://24-ai.news/en/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI en 2026-06-20T00:00:00Z AMD: ROCm Optimization of Matrix3D for 3D Worlds Speeds Up Rendering by up to 54 Percent on Instinct GPUs https://24-ai.news/en/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI en 2026-06-20T00:00:00Z Anthropic: Claude Code v2.1.183 Blocks Destructive Git and Infrastructure Commands in Auto Mode https://24-ai.news/en/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI en 2026-06-20T00:00:00Z arXiv:2606.20205: Psychological Profiles of LLMs Are Largely a Measurement Artifact, Not a Stable Personality https://24-ai.news/en/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI en 2026-06-20T00:00:00Z arXiv:2606.20508: What Language Models Learn from Mixed Demonstrations of Safe and Harmful Behavior https://24-ai.news/en/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI en 2026-06-20T00:00:00Z arXiv:2606.20517: Multi-LCB Extends LiveCodeBench to 12 Programming Languages and Reveals Python Overfitting in 24 Models https://24-ai.news/en/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI en 2026-06-20T00:00:00Z arXiv:2606.19808: SEVRA Saves up to 91 Percent of Tokens Through Selective Verification in Model Reasoning https://24-ai.news/en/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI en 2026-06-20T00:00:00Z arXiv:2606.20333: SoftSkill Compresses Skill Documents into 32 Latent Tokens and Boosts LiveMath by 42.1 Points https://24-ai.news/en/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI en 2026-06-20T00:00:00Z AWS: SageMaker Gets Over 100 Detailed Inference Metrics and an Insights Dashboard on CloudWatch https://24-ai.news/en/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI en 2026-06-20T00:00:00Z EU AI Office: The EUROPA Consortium Is Building an Open European Frontier Model with Over 400 Billion Parameters for All 24 EU Languages https://24-ai.news/en/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI en 2026-06-20T00:00:00Z GitHub: Copilot Retires Opus 4.6 (fast) on June 29, Adds AGENTS.md to Code Review and ai_credits_used Field to API https://24-ai.news/en/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI en 2026-06-20T00:00:00Z GitHub: Internal Analytics Agent Qubot Reduced Query Resolution Time by Around 66 Percent https://24-ai.news/en/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI en 2026-06-20T00:00:00Z NVIDIA: Partners at Cannes Lions 2026 Showcased Agentic AI for Marketing, Criteo Doubled Training Speed on Blackwell https://24-ai.news/en/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI en 2026-06-20T00:00:00Z UK AISI: Engineering Playbook Opens Frontier Model Evaluation Infrastructure in Five Layers https://24-ai.news/de/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI de 2026-06-20T00:00:00Z AMD: ROCm-Optimierung von Matrix3D für 3D-Welten beschleunigt Rendering auf Instinct-GPUs um bis zu 54 Prozent https://24-ai.news/de/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI de 2026-06-20T00:00:00Z Anthropic: Claude Code v2.1.183 blockiert destruktive Git- und Infrastrukturbefehle im Auto-Modus https://24-ai.news/de/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI de 2026-06-20T00:00:00Z arXiv:2606.20205: Psychologische Profile von Sprachmodellen sind größtenteils ein Messartefakt, keine stabile Persönlichkeit https://24-ai.news/de/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI de 2026-06-20T00:00:00Z arXiv:2606.20508: Was Sprachmodelle aus gemischten Demonstrations sicheren und schädlichen Verhaltens lernen https://24-ai.news/de/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI de 2026-06-20T00:00:00Z arXiv:2606.20517: Multi-LCB erweitert LiveCodeBench auf 12 Programmiersprachen und deckt Python-Overfitting bei 24 Modellen auf https://24-ai.news/de/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI de 2026-06-20T00:00:00Z arXiv:2606.19808: SEVRA spart durch selektive Verifikation bis zu 91 Prozent der Tokens beim Modell-Reasoning https://24-ai.news/de/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI de 2026-06-20T00:00:00Z arXiv:2606.20333: SoftSkill komprimiert Skill-Dokumente in 32 latente Tokens und steigert LiveMath um 42,1 Punkte https://24-ai.news/de/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI de 2026-06-20T00:00:00Z AWS: SageMaker erhält über 100 detaillierte Inference-Metriken und Insights-Dashboard auf CloudWatch https://24-ai.news/de/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI de 2026-06-20T00:00:00Z EU AI Office: Konsortium EUROPA baut offenes europäisches Frontier-Modell mit über 400 Milliarden Parametern für alle 24 EU-Sprachen https://24-ai.news/de/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI de 2026-06-20T00:00:00Z GitHub: Copilot stellt Opus 4.6 (fast) am 29. Juni ein, fügt AGENTS.md zu Code Review hinzu und ai_credits_used-Feld zur API https://24-ai.news/de/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI de 2026-06-20T00:00:00Z GitHub: Interner Analyse-Agent Qubot verkürzte Abfragezeit um rund 66 Prozent https://24-ai.news/de/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI de 2026-06-20T00:00:00Z NVIDIA: Partner auf Cannes Lions 2026 zeigten agentische KI für Marketing, Criteo verdoppelte Trainingsgeschwindigkeit auf Blackwell https://24-ai.news/de/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI de 2026-06-20T00:00:00Z UK AISI: Engineering Playbook öffnet Frontier-Modell-Evaluierungsinfrastruktur in fünf Schichten https://24-ai.news/zh/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI zh 2026-06-20T00:00:00Z AMD：ROCm优化Matrix3D框架，在Instinct GPU上将3D世界渲染速度提升最高54% https://24-ai.news/zh/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI zh 2026-06-20T00:00:00Z Anthropic：Claude Code v2.1.183在自动模式下屏蔽破坏性git和基础设施命令 https://24-ai.news/zh/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI zh 2026-06-20T00:00:00Z arXiv:2606.20205：语言模型的心理特征画像主要是测量偏差，而非稳定人格 https://24-ai.news/zh/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI zh 2026-06-20T00:00:00Z arXiv:2606.20508：语言模型从混合无害与有害示范中学到了什么 https://24-ai.news/zh/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI zh 2026-06-20T00:00:00Z arXiv:2606.20517：Multi-LCB将LiveCodeBench扩展至12种编程语言，揭示24个模型的Python过拟合 https://24-ai.news/zh/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI zh 2026-06-20T00:00:00Z arXiv:2606.19808：SEVRA通过选择性验证可节省模型推理最多91%的token https://24-ai.news/zh/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI zh 2026-06-20T00:00:00Z arXiv:2606.20333：SoftSkill将技能文档压缩为32个潜在token，LiveMath提升42.1分 https://24-ai.news/zh/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI zh 2026-06-20T00:00:00Z AWS：SageMaker获得100余项详细推理指标和CloudWatch Insights监控面板 https://24-ai.news/zh/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI zh 2026-06-20T00:00:00Z EU AI Office：EUROPA联盟建设支持24种欧盟语言的4000亿参数开放欧洲前沿模型 https://24-ai.news/zh/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI zh 2026-06-20T00:00:00Z GitHub：Copilot于6月29日停用Opus 4.6 (fast)，code review新增AGENTS.md，API增加ai_credits_used字段 https://24-ai.news/zh/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI zh 2026-06-20T00:00:00Z GitHub：内部分析智能体Qubot将查询解决时间缩短约66% https://24-ai.news/zh/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI zh 2026-06-20T00:00:00Z NVIDIA：合作伙伴在2026年戛纳创意节展示营销智能体AI，Criteo在Blackwell上将训练速度提升2倍 https://24-ai.news/zh/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI zh 2026-06-20T00:00:00Z UK AISI：Engineering Playbook以五层结构开放前沿模型评估基础设施 https://24-ai.news/ja/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI ja 2026-06-20T00:00:00Z AMD：ROCm最適化でMatrix3DのInstinct GPU向け3D世界生成が最大54%高速化 https://24-ai.news/ja/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI ja 2026-06-20T00:00:00Z Anthropic：Claude Code v2.1.183がオートモードで破壊的なgit・インフラコマンドをブロック https://24-ai.news/ja/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI ja 2026-06-20T00:00:00Z arXiv:2606.20205：LLMの心理プロファイルは安定した人格ではなく測定アーティファクトである https://24-ai.news/ja/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI ja 2026-06-20T00:00:00Z arXiv:2606.20508：LLMは無害・有害デモンストレーションの混合から何を学ぶか https://24-ai.news/ja/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI ja 2026-06-20T00:00:00Z arXiv:2606.20517：Multi-LCBがLiveCodeBenchを12言語に拡張、24モデルのPython過学習を発見 https://24-ai.news/ja/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI ja 2026-06-20T00:00:00Z arXiv:2606.19808：SEVRAの選択的検証でモデル推論のトークンを最大91%削減 https://24-ai.news/ja/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI ja 2026-06-20T00:00:00Z arXiv:2606.20333：SoftSkillがスキル文書を32の潜在トークンに圧縮、LiveMathを42.1ポイント向上 https://24-ai.news/ja/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI ja 2026-06-20T00:00:00Z AWS：SageMakerが100以上の詳細な推論メトリクスとCloudWatch Insightsダッシュボードを取得 https://24-ai.news/ja/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI ja 2026-06-20T00:00:00Z EU AI Office：EUROPAコンソーシアム、EU24言語対応の4000億パラメータ欧州オープンフロンティアモデルを構築へ https://24-ai.news/ja/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI ja 2026-06-20T00:00:00Z GitHub：CopilotがOpus 4.6 (fast)を6月29日に廃止、code reviewにAGENTS.md追加、APIにai_credits_usedフィールド https://24-ai.news/ja/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI ja 2026-06-20T00:00:00Z GitHub：内部分析エージェント「Qubot」がクエリ解決時間を約66%短縮 https://24-ai.news/ja/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI ja 2026-06-20T00:00:00Z NVIDIA：カンヌライオンズ2026でパートナーがマーケティング向けエージェントAIを披露、CriteoはBlackwellでトレーニング速度2倍 https://24-ai.news/ja/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI ja 2026-06-20T00:00:00Z UK AISI：Engineering Playbookでフロンティアモデル評価インフラを5層構造で公開 https://24-ai.news/ko/news/2026-06-20/amd-rocm-matrix3d-world-generation/ 24 AI ko 2026-06-20T00:00:00Z AMD：ROCm으로 Matrix3D 최적화, Instinct GPU에서 3D 세계 렌더링 최대 54% 가속 https://24-ai.news/ko/news/2026-06-20/anthropic-claude-code-2-1-183-safety/ 24 AI ko 2026-06-20T00:00:00Z Anthropic：Claude Code v2.1.183, 자동 모드에서 파괴적 git·인프라 명령 차단 https://24-ai.news/ko/news/2026-06-20/arxiv-llm-psychology-profiles-artifact/ 24 AI ko 2026-06-20T00:00:00Z arXiv:2606.20205：LLM의 심리 프로필은 안정된 인격이 아닌 측정 아티팩트 https://24-ai.news/ko/news/2026-06-20/arxiv-mixed-compliance-demonstrations/ 24 AI ko 2026-06-20T00:00:00Z arXiv:2606.20508：언어 모델은 무해·유해 데모의 혼합에서 무엇을 학습하는가 https://24-ai.news/ko/news/2026-06-20/arxiv-multi-lcb-livecodebench-12-languages/ 24 AI ko 2026-06-20T00:00:00Z arXiv:2606.20517：Multi-LCB, LiveCodeBench를 12개 프로그래밍 언어로 확장하고 24개 모델의 Python 과적합 발견 https://24-ai.news/ko/news/2026-06-20/arxiv-sevra-selective-verification/ 24 AI ko 2026-06-20T00:00:00Z arXiv:2606.19808：SEVRA, 선택적 검증으로 모델 추론 토큰 최대 91% 절감 https://24-ai.news/ko/news/2026-06-20/arxiv-softskill-behavioral-compression/ 24 AI ko 2026-06-20T00:00:00Z arXiv:2606.20333：SoftSkill이 스킬 문서를 32개 잠재 토큰으로 압축, LiveMath 42.1점 향상 https://24-ai.news/ko/news/2026-06-20/aws-sagemaker-inference-metrics-dashboard/ 24 AI ko 2026-06-20T00:00:00Z AWS：SageMaker에 100개 이상의 상세 추론 지표와 CloudWatch Insights 대시보드 추가 https://24-ai.news/ko/news/2026-06-20/eu-ai-office-europa-frontier-challenge/ 24 AI ko 2026-06-20T00:00:00Z EU AI Office：EUROPA 컨소시엄, EU 24개 언어 지원 4000억 파라미터 오픈 유럽 프런티어 모델 구축 https://24-ai.news/ko/news/2026-06-20/github-copilot-changelog-opus-46-sunset/ 24 AI ko 2026-06-20T00:00:00Z GitHub：Copilot이 6월 29일 Opus 4.6 (fast) 종료, code review에 AGENTS.md 추가, API에 ai_credits_used 필드 https://24-ai.news/ko/news/2026-06-20/github-qubot-data-analytics-agent/ 24 AI ko 2026-06-20T00:00:00Z GitHub：내부 분석 에이전트 Qubot이 쿼리 해결 시간 약 66% 단축 https://24-ai.news/ko/news/2026-06-20/nvidia-agentic-marketing-cannes-lions/ 24 AI ko 2026-06-20T00:00:00Z NVIDIA：2026 칸 라이언스에서 파트너들이 마케팅용 에이전트 AI 공개, Criteo는 Blackwell에서 훈련 속도 2배 https://24-ai.news/ko/news/2026-06-20/uk-aisi-engineering-playbook/ 24 AI ko 2026-06-20T00:00:00Z UK AISI：Engineering Playbook으로 프런티어 모델 평가 인프라를 5개 레이어로 공개 https://24-ai.news/hr/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI hr 2026-06-19T00:00:00Z AMD: Analiza RoCE mrežnih uzoraka pri treningu velikih jezičnih modela https://24-ai.news/hr/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI hr 2026-06-19T00:00:00Z Anthropic: Project Fetch — faza dva pokazuje 20× brži robotizirani rad uz 10× manje koda https://24-ai.news/hr/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI hr 2026-06-19T00:00:00Z arXiv:2606.18543: CEO-Bench — mogu li agenti voditi startup na dugi rok? https://24-ai.news/hr/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI hr 2026-06-19T00:00:00Z arXiv:2606.19327: Rubric-conditioned self-distillation nadmašuje GRPO u treniranju reasoning modela https://24-ai.news/hr/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI hr 2026-06-19T00:00:00Z arXiv:2606.19245: TxBench-PP — AI agenti u potrazi za novim lijekovima https://24-ai.news/hr/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI hr 2026-06-19T00:00:00Z AWS: Amazon Bedrock AgentCore Harness dostigao opću dostupnost — deploy u 2 API poziva https://24-ai.news/hr/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI hr 2026-06-19T00:00:00Z Black Forest Labs: Robin Rombach poziva G7 lidere na podršku otvorenom razvoju AI-a https://24-ai.news/hr/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI hr 2026-06-19T00:00:00Z Google DeepMind: Više od 50% sigurnosnih incidenata agenata su greške, ne napadi https://24-ai.news/hr/news/2026-06-19/github-actions-security-pwn-request/ 24 AI hr 2026-06-19T00:00:00Z GitHub: Dvije sigurnosne nadogradnje GitHub Actions štite od pwn request napada https://24-ai.news/hr/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI hr 2026-06-19T00:00:00Z GitHub: AI agenti sada mogu kreirati potpuno triagirane issue-e bez ljudske intervencije https://24-ai.news/hr/news/2026-06-19/github-mai-code-1-flash/ 24 AI hr 2026-06-19T00:00:00Z GitHub: MAI-Code-1-Flash dostupan na 8 razvojnih površina unutar Copilota https://24-ai.news/hr/news/2026-06-19/github-pull-request-limits/ 24 AI hr 2026-06-19T00:00:00Z GitHub: Kako PR limiti smanjuju buku u open source projektima https://24-ai.news/hr/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI hr 2026-06-19T00:00:00Z OpenAI: GPT-5.5 Instant donosi napredak u zdravstvenim odgovorima ChatGPT-a https://24-ai.news/hr/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI hr 2026-06-19T00:00:00Z OpenAI: AI pomaže liječnicima dijagnosticirati rijetke genetske bolesti kod djece https://24-ai.news/hr/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI hr 2026-06-19T00:00:00Z PyTorch: LLM-ovi smanjuju optimizaciju GPU kernela s minuta na sekunde https://24-ai.news/en/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI en 2026-06-19T00:00:00Z AMD: Analysis of RoCE Network Traffic Patterns in Large Language Model Training https://24-ai.news/en/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI en 2026-06-19T00:00:00Z Anthropic: Project Fetch Phase Two Shows 20× Faster Robotic Operation with 10× Less Code https://24-ai.news/en/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI en 2026-06-19T00:00:00Z arXiv:2606.18543: CEO-Bench — Can Agents Run a Startup for the Long Term? https://24-ai.news/en/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI en 2026-06-19T00:00:00Z arXiv:2606.19327: Rubric-Conditioned Self-Distillation Outperforms GRPO in Reasoning Model Training https://24-ai.news/en/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI en 2026-06-19T00:00:00Z arXiv:2606.19245: TxBench-PP — AI Agents in the Search for New Drugs https://24-ai.news/en/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI en 2026-06-19T00:00:00Z AWS: Amazon Bedrock AgentCore Harness Reaches General Availability — Deploy in 2 API Calls https://24-ai.news/en/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI en 2026-06-19T00:00:00Z Black Forest Labs: Robin Rombach Calls on G7 Leaders to Support Open AI Development https://24-ai.news/en/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI en 2026-06-19T00:00:00Z Google DeepMind: Over 50% of Agent Security Incidents Are Mistakes, Not Attacks https://24-ai.news/en/news/2026-06-19/github-actions-security-pwn-request/ 24 AI en 2026-06-19T00:00:00Z GitHub: Two Security Upgrades to GitHub Actions Protect Against Pwn Request Attacks https://24-ai.news/en/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI en 2026-06-19T00:00:00Z GitHub: AI Agents Can Now Create Fully Triaged Issues Without Human Intervention https://24-ai.news/en/news/2026-06-19/github-mai-code-1-flash/ 24 AI en 2026-06-19T00:00:00Z GitHub: MAI-Code-1-Flash Available on 8 Development Surfaces Within Copilot https://24-ai.news/en/news/2026-06-19/github-pull-request-limits/ 24 AI en 2026-06-19T00:00:00Z GitHub: How Pull Request Limits Are Reducing Noise in Open Source Projects https://24-ai.news/en/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI en 2026-06-19T00:00:00Z OpenAI: GPT-5.5 Instant Advances Health Intelligence in ChatGPT https://24-ai.news/en/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI en 2026-06-19T00:00:00Z OpenAI: AI Helps Physicians Diagnose Rare Genetic Diseases in Children https://24-ai.news/en/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI en 2026-06-19T00:00:00Z PyTorch: LLMs Reduce GPU Kernel Optimization from Minutes to Seconds https://24-ai.news/de/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI de 2026-06-19T00:00:00Z AMD: Analyse von RoCE-Netzwerkverkehrsmustern beim Training großer Sprachmodelle https://24-ai.news/de/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI de 2026-06-19T00:00:00Z Anthropic: Project Fetch Phase 2 — 20× schnellerer Roboterbetrieb mit 10× weniger Code https://24-ai.news/de/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI de 2026-06-19T00:00:00Z arXiv:2606.18543: CEO-Bench — Können KI-Agenten ein Startup langfristig führen? https://24-ai.news/de/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI de 2026-06-19T00:00:00Z arXiv:2606.19327: Rubrik-konditionierte Selbstdestillation übertrifft GRPO beim Reasoning-Training https://24-ai.news/de/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI de 2026-06-19T00:00:00Z arXiv:2606.19245: TxBench-PP — KI-Agenten auf der Suche nach neuen Medikamenten https://24-ai.news/de/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI de 2026-06-19T00:00:00Z AWS: Amazon Bedrock AgentCore Harness erreicht allgemeine Verfügbarkeit — Deploy in 2 API-Aufrufen https://24-ai.news/de/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI de 2026-06-19T00:00:00Z Black Forest Labs: Robin Rombach fordert G7-Führer auf, offene KI-Entwicklung zu unterstützen https://24-ai.news/de/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI de 2026-06-19T00:00:00Z Google DeepMind: Über 50 % der KI-Agenten-Sicherheitsvorfälle sind Fehler, keine Angriffe https://24-ai.news/de/news/2026-06-19/github-actions-security-pwn-request/ 24 AI de 2026-06-19T00:00:00Z GitHub: Zwei Sicherheitsupdates für GitHub Actions schützen vor Pwn-Request-Angriffen https://24-ai.news/de/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI de 2026-06-19T00:00:00Z GitHub: KI-Agenten können nun vollständig triagierte Issues ohne menschliche Eingriffe erstellen https://24-ai.news/de/news/2026-06-19/github-mai-code-1-flash/ 24 AI de 2026-06-19T00:00:00Z GitHub: MAI-Code-1-Flash auf 8 Entwicklungsoberflächen innerhalb von Copilot verfügbar https://24-ai.news/de/news/2026-06-19/github-pull-request-limits/ 24 AI de 2026-06-19T00:00:00Z GitHub: Wie Pull-Request-Limits den Lärm in Open-Source-Projekten reduzieren https://24-ai.news/de/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI de 2026-06-19T00:00:00Z OpenAI: GPT-5.5 Instant verbessert Gesundheitsintelligenz in ChatGPT https://24-ai.news/de/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI de 2026-06-19T00:00:00Z OpenAI: KI hilft Ärzten bei der Diagnose seltener genetischer Erkrankungen bei Kindern https://24-ai.news/de/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI de 2026-06-19T00:00:00Z PyTorch: LLMs reduzieren GPU-Kernel-Optimierung von Minuten auf Sekunden https://24-ai.news/zh/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI zh 2026-06-19T00:00:00Z AMD：大型语言模型训练中RoCE网络流量模式分析 https://24-ai.news/zh/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI zh 2026-06-19T00:00:00Z Anthropic：Project Fetch第二阶段——机器人作业速度提升20倍，代码量减少10倍 https://24-ai.news/zh/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI zh 2026-06-19T00:00:00Z arXiv:2606.18543：CEO-Bench——AI智能体能长期经营初创企业吗？ https://24-ai.news/zh/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI zh 2026-06-19T00:00:00Z arXiv:2606.19327：基于评分标准的自蒸馏在推理模型训练中超越GRPO https://24-ai.news/zh/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI zh 2026-06-19T00:00:00Z arXiv:2606.19245：TxBench-PP——AI智能体探索新药研发 https://24-ai.news/zh/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI zh 2026-06-19T00:00:00Z AWS：Amazon Bedrock AgentCore Harness正式发布——仅需2次API调用完成部署 https://24-ai.news/zh/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI zh 2026-06-19T00:00:00Z Black Forest Labs：Robin Rombach呼吁G7领导人支持开放式AI发展 https://24-ai.news/zh/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI zh 2026-06-19T00:00:00Z Google DeepMind：超过50%的AI智能体安全事件源于错误，而非攻击 https://24-ai.news/zh/news/2026-06-19/github-actions-security-pwn-request/ 24 AI zh 2026-06-19T00:00:00Z GitHub：两项安全升级保护GitHub Actions免受pwn request攻击 https://24-ai.news/zh/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI zh 2026-06-19T00:00:00Z GitHub：AI智能体现可无需人工干预创建完整分类的issue https://24-ai.news/zh/news/2026-06-19/github-mai-code-1-flash/ 24 AI zh 2026-06-19T00:00:00Z GitHub：MAI-Code-1-Flash现已在Copilot的8个开发界面上提供 https://24-ai.news/zh/news/2026-06-19/github-pull-request-limits/ 24 AI zh 2026-06-19T00:00:00Z GitHub：PR限制如何减少开源项目中的噪音 https://24-ai.news/zh/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI zh 2026-06-19T00:00:00Z OpenAI：GPT-5.5 Instant提升ChatGPT健康领域回答质量 https://24-ai.news/zh/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI zh 2026-06-19T00:00:00Z OpenAI：AI帮助医生诊断儿童罕见遗传病 https://24-ai.news/zh/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI zh 2026-06-19T00:00:00Z PyTorch：LLM将GPU内核优化时间从分钟缩短至秒级 https://24-ai.news/ja/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI ja 2026-06-19T00:00:00Z AMD：大規模言語モデル訓練におけるRoCEネットワークトラフィックパターン分析 https://24-ai.news/ja/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI ja 2026-06-19T00:00:00Z Anthropic：Project Fetch第2フェーズ——ロボット作業が20倍高速化、コード量10倍削減 https://24-ai.news/ja/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI ja 2026-06-19T00:00:00Z arXiv:2606.18543：CEO-Bench——AIエージェントは長期にわたってスタートアップを経営できるか？ https://24-ai.news/ja/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI ja 2026-06-19T00:00:00Z arXiv:2606.19327：ルーブリック条件付き自己蒸留が推論モデル訓練でGRPOを凌駕 https://24-ai.news/ja/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI ja 2026-06-19T00:00:00Z arXiv:2606.19245：TxBench-PP——AIエージェントが新薬探索に挑む https://24-ai.news/ja/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI ja 2026-06-19T00:00:00Z AWS：Amazon Bedrock AgentCore Harness、一般提供開始——2回のAPI呼び出しでデプロイ完了 https://24-ai.news/ja/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI ja 2026-06-19T00:00:00Z Black Forest Labs：Robin RombachがG7リーダーにオープンAI開発への支持を訴える https://24-ai.news/ja/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI ja 2026-06-19T00:00:00Z Google DeepMind：AIエージェントのセキュリティインシデントの50%超は攻撃ではなくエラーが原因 https://24-ai.news/ja/news/2026-06-19/github-actions-security-pwn-request/ 24 AI ja 2026-06-19T00:00:00Z GitHub：2つのセキュリティアップデートがGitHub Actionsをpwnリクエスト攻撃から守る https://24-ai.news/ja/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI ja 2026-06-19T00:00:00Z GitHub：AIエージェントが人間の介入なしに完全にトリアージされたissueを作成できるようになった https://24-ai.news/ja/news/2026-06-19/github-mai-code-1-flash/ 24 AI ja 2026-06-19T00:00:00Z GitHub：MAI-Code-1-FlashがCopilotの8つの開発サーフェスで利用可能に https://24-ai.news/ja/news/2026-06-19/github-pull-request-limits/ 24 AI ja 2026-06-19T00:00:00Z GitHub：PRリミットがオープンソースプロジェクトのノイズをどのように削減するか https://24-ai.news/ja/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI ja 2026-06-19T00:00:00Z OpenAI：GPT-5.5 InstantがChatGPTの健康分野における回答を改善 https://24-ai.news/ja/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI ja 2026-06-19T00:00:00Z OpenAI：AIが小児の希少遺伝疾患の診断を支援 https://24-ai.news/ja/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI ja 2026-06-19T00:00:00Z PyTorch：LLMがGPUカーネル最適化を分単位から秒単位に短縮 https://24-ai.news/ko/news/2026-06-19/amd-roce-llm-training-analysis/ 24 AI ko 2026-06-19T00:00:00Z AMD: 대규모 언어 모델 학습에서의 RoCE 네트워크 트래픽 패턴 분석 https://24-ai.news/ko/news/2026-06-19/anthropic-project-fetch-phase-two/ 24 AI ko 2026-06-19T00:00:00Z Anthropic: Project Fetch 2단계——로봇 작업 20배 빠르고 코드 10배 감소 https://24-ai.news/ko/news/2026-06-19/arxiv-ceo-bench-agent-startup/ 24 AI ko 2026-06-19T00:00:00Z arXiv:2606.18543: CEO-Bench——AI 에이전트가 장기적으로 스타트업을 이끌 수 있을까? https://24-ai.news/ko/news/2026-06-19/arxiv-rubric-self-distillation/ 24 AI ko 2026-06-19T00:00:00Z arXiv:2606.19327: 루브릭 조건부 자기 증류, 추론 모델 학습에서 GRPO 초월 https://24-ai.news/ko/news/2026-06-19/arxiv-txbench-pp-drug-discovery/ 24 AI ko 2026-06-19T00:00:00Z arXiv:2606.19245: TxBench-PP——AI 에이전트, 신약 개발에 도전하다 https://24-ai.news/ko/news/2026-06-19/aws-bedrock-agentcore-harness-ga/ 24 AI ko 2026-06-19T00:00:00Z AWS: Amazon Bedrock AgentCore Harness 일반 공개——2번의 API 호출로 배포 완료 https://24-ai.news/ko/news/2026-06-19/black-forest-labs-g7-open-innovation/ 24 AI ko 2026-06-19T00:00:00Z Black Forest Labs: Robin Rombach, G7 지도자들에게 개방형 AI 개발 지지 촉구 https://24-ai.news/ko/news/2026-06-19/deepmind-securing-ai-agents/ 24 AI ko 2026-06-19T00:00:00Z Google DeepMind: AI 에이전트 보안 사고의 50% 이상은 공격이 아닌 오류가 원인 https://24-ai.news/ko/news/2026-06-19/github-actions-security-pwn-request/ 24 AI ko 2026-06-19T00:00:00Z GitHub: 두 가지 보안 업데이트로 GitHub Actions를 pwn request 공격으로부터 보호 https://24-ai.news/ko/news/2026-06-19/github-issue-duplicate-detection-mcp/ 24 AI ko 2026-06-19T00:00:00Z GitHub: AI 에이전트가 이제 인간 개입 없이 완전히 분류된 이슈를 생성할 수 있다 https://24-ai.news/ko/news/2026-06-19/github-mai-code-1-flash/ 24 AI ko 2026-06-19T00:00:00Z GitHub: MAI-Code-1-Flash, Copilot의 8개 개발 인터페이스에서 이용 가능 https://24-ai.news/ko/news/2026-06-19/github-pull-request-limits/ 24 AI ko 2026-06-19T00:00:00Z GitHub: PR 제한이 오픈소스 프로젝트의 노이즈를 줄이는 방법 https://24-ai.news/ko/news/2026-06-19/openai-gpt-5-5-health-intelligence/ 24 AI ko 2026-06-19T00:00:00Z OpenAI: GPT-5.5 Instant, ChatGPT 건강 분야 답변 개선 https://24-ai.news/ko/news/2026-06-19/openai-rare-disease-diagnosis-children/ 24 AI ko 2026-06-19T00:00:00Z OpenAI: AI가 의사들의 아동 희귀 유전 질환 진단을 돕다 https://24-ai.news/ko/news/2026-06-19/pytorch-helion-llm-autotuning/ 24 AI ko 2026-06-19T00:00:00Z PyTorch: LLM이 GPU 커널 최적화 시간을 분에서 초로 단축 https://24-ai.news/hr/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI hr 2026-06-18T00:00:00Z Allen Institute: Open-source MolmoMotion predviđa 3D kretanje iz videa i postavlja SOTA u robotici https://24-ai.news/hr/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI hr 2026-06-18T00:00:00Z AMD: Open-source Schola spaja Unreal Engine i učenje pojačanjem za trening robotske ruke na ROCm-u https://24-ai.news/hr/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI hr 2026-06-18T00:00:00Z Anthropic: Claude Code v2.1.181 uvodi /config sintaksu i Bun 1.4 te auto-retry pri padu API-ja https://24-ai.news/hr/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI hr 2026-06-18T00:00:00Z Anthropic: Otvoren ured u Seulu i nova partnerstva u korejskom AI ekosustavu https://24-ai.news/hr/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI hr 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens otkriva da 52% prosjek halucinacija u pravnom AI-ju krije 38 bodova razlike https://24-ai.news/hr/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI hr 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBench pokazuje da agentski AI širi pseudoznanost uz gotovo nultu stopu odbijanja https://24-ai.news/hr/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI hr 2026-06-18T00:00:00Z AWS: Bedrock AgentCore dobiva web pretragu, plaćanja i A/B testiranje agenata https://24-ai.news/hr/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI hr 2026-06-18T00:00:00Z CNCF: Kubernetes kao operativna osnova agentskog AI-ja — pouke iz multi-agent security platforme https://24-ai.news/hr/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI hr 2026-06-18T00:00:00Z EU: Izvještaj o stanju digitalnog desetljeća 2026 traži zatvaranje strukturnih jazova do 2030. https://24-ai.news/hr/news/2026-06-18/github-copilot-app-ga/ 24 AI hr 2026-06-18T00:00:00Z GitHub: Copilot desktop aplikacija u općoj dostupnosti — paralelne sesije i cloud automacije https://24-ai.news/hr/news/2026-06-18/github-copilot-hydra-router/ 24 AI hr 2026-06-18T00:00:00Z GitHub: Copilot HyDRA router automatski bira model i donosi 72,5% uštede uz istu kvalitetu https://24-ai.news/hr/news/2026-06-18/google-amie-primary-care-nature/ 24 AI hr 2026-06-18T00:00:00Z Google: AMIE u Nature studiji dostigao razinu liječnika obiteljske medicine u vođenju bolesti https://24-ai.news/hr/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI hr 2026-06-18T00:00:00Z IBM: Studija otkriva da 91% organizacija nema punu vidljivost AI ovisnosti, lock-in ozbiljan https://24-ai.news/hr/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI hr 2026-06-18T00:00:00Z OpenAI: Gotovo autonomni AI kemičar temeljen na GPT-5.4 poboljšao reakciju u sintezi lijekova https://24-ai.news/en/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI en 2026-06-18T00:00:00Z Allen Institute: Open-Source MolmoMotion Predicts 3D Motion from Video and Sets SOTA in Robotics https://24-ai.news/en/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI en 2026-06-18T00:00:00Z AMD: Open-source Schola connects Unreal Engine and reinforcement learning for robotic arm training on ROCm https://24-ai.news/en/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI en 2026-06-18T00:00:00Z Anthropic: Claude Code v2.1.181 introduces /config syntax and Bun 1.4 with auto-retry on API failures https://24-ai.news/en/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI en 2026-06-18T00:00:00Z Anthropic: Seoul office opens and new partnerships announced in the Korean AI ecosystem https://24-ai.news/en/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI en 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens reveals that a 52% hallucination average in legal AI hides a 38-point gap https://24-ai.news/en/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI en 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBench Shows Agentic AI Spreads Pseudoscience with Near-Zero Rejection Rate https://24-ai.news/en/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI en 2026-06-18T00:00:00Z AWS: Bedrock AgentCore Gets Web Search, Payments, and A/B Testing for Agents https://24-ai.news/en/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI en 2026-06-18T00:00:00Z CNCF: Kubernetes as the Operational Foundation for Agentic AI — Lessons from a Multi-Agent Security Platform https://24-ai.news/en/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI en 2026-06-18T00:00:00Z EU: 2026 State of the Digital Decade Report calls for closing structural gaps by 2030. https://24-ai.news/en/news/2026-06-18/github-copilot-app-ga/ 24 AI en 2026-06-18T00:00:00Z GitHub: Copilot desktop app now generally available — parallel sessions and cloud automations https://24-ai.news/en/news/2026-06-18/github-copilot-hydra-router/ 24 AI en 2026-06-18T00:00:00Z GitHub: Copilot HyDRA Router Automatically Selects Model and Delivers 72.5% Savings at Same Quality https://24-ai.news/en/news/2026-06-18/google-amie-primary-care-nature/ 24 AI en 2026-06-18T00:00:00Z Google: AMIE Reached Primary Care Physician Level in Nature Study on Disease Management https://24-ai.news/en/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI en 2026-06-18T00:00:00Z IBM: Study Reveals 91% of Organizations Lack Full Visibility into AI Dependencies, Lock-In Serious https://24-ai.news/en/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI en 2026-06-18T00:00:00Z OpenAI: Near-Autonomous AI Chemist Based on GPT-5.4 Improves Reaction in Drug Synthesis https://24-ai.news/de/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI de 2026-06-18T00:00:00Z Allen Institute: Open-Source MolmoMotion sagt 3D-Bewegung aus Videos vorher und setzt SOTA in der Robotik https://24-ai.news/de/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI de 2026-06-18T00:00:00Z AMD: Open-source Schola verbindet Unreal Engine und Reinforcement Learning für Roboterarm-Training auf ROCm https://24-ai.news/de/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI de 2026-06-18T00:00:00Z Anthropic: Claude Code v2.1.181 führt /config-Syntax und Bun 1.4 sowie Auto-Retry bei API-Ausfällen ein https://24-ai.news/de/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI de 2026-06-18T00:00:00Z Anthropic: Büro in Seoul eröffnet und neue Partnerschaften im koreanischen KI-Ökosystem angekündigt https://24-ai.news/de/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI de 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens zeigt, dass 52 % Halluzinationsdurchschnitt im Rechts-KI 38 Punkte Unterschied verbirgt https://24-ai.news/de/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI de 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBench zeigt, dass agentische KI Pseudowissenschaft mit nahezu nullprozentiger Ablehnungsrate verbreitet https://24-ai.news/de/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI de 2026-06-18T00:00:00Z AWS: Bedrock AgentCore erhält Websuche, Zahlungen und A/B-Tests für Agenten https://24-ai.news/de/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI de 2026-06-18T00:00:00Z CNCF: Kubernetes als Betriebsgrundlage für agentische KI — Lehren aus einer Multi-Agenten-Sicherheitsplattform https://24-ai.news/de/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI de 2026-06-18T00:00:00Z EU: Bericht zum Stand des Digitalen Jahrzehnts 2026 fordert Schließung struktureller Lücken bis 2030. https://24-ai.news/de/news/2026-06-18/github-copilot-app-ga/ 24 AI de 2026-06-18T00:00:00Z GitHub: Copilot-Desktop-App allgemein verfügbar — parallele Sitzungen und Cloud-Automatisierungen https://24-ai.news/de/news/2026-06-18/github-copilot-hydra-router/ 24 AI de 2026-06-18T00:00:00Z GitHub: Copilot HyDRA-Router wählt Modell automatisch und liefert 72,5 % Ersparnis bei gleicher Qualität https://24-ai.news/de/news/2026-06-18/google-amie-primary-care-nature/ 24 AI de 2026-06-18T00:00:00Z Google: AMIE erreicht in Nature-Studie das Niveau von Hausärzten im Krankheitsmanagement https://24-ai.news/de/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI de 2026-06-18T00:00:00Z IBM: Studie zeigt, dass 91 % der Unternehmen keine vollständige Sicht auf KI-Abhängigkeiten haben, Lock-in ernst https://24-ai.news/de/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI de 2026-06-18T00:00:00Z OpenAI: Fast autonomer KI-Chemiker auf GPT-5.4-Basis verbessert Reaktion in der Wirkstoffsynthese https://24-ai.news/zh/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI zh 2026-06-18T00:00:00Z Allen Institute：开源MolmoMotion从视频中预测3D运动，在机器人领域达到SOTA https://24-ai.news/zh/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI zh 2026-06-18T00:00:00Z AMD：开源 Schola 在 ROCm 平台上连接 Unreal Engine 与强化学习，用于机器人手臂训练 https://24-ai.news/zh/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI zh 2026-06-18T00:00:00Z Anthropic：Claude Code v2.1.181 引入 /config 语法与 Bun 1.4，并在 API 连接失败时自动重试 https://24-ai.news/zh/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI zh 2026-06-18T00:00:00Z Anthropic：首尔办公室开业，并宣布与韩国AI生态系统建立新合作伙伴关系 https://24-ai.news/zh/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI zh 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens 揭示法律 AI 52% 幻觉平均值背后隐藏着 38 个百分点的差距 https://24-ai.news/zh/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI zh 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBench揭示智能体AI传播伪科学的拒绝率接近于零 https://24-ai.news/zh/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI zh 2026-06-18T00:00:00Z AWS：Bedrock AgentCore新增网络搜索、支付功能和代理A/B测试 https://24-ai.news/zh/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI zh 2026-06-18T00:00:00Z CNCF：Kubernetes作为智能体AI的运营基础——多智能体安全平台的经验教训 https://24-ai.news/zh/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI zh 2026-06-18T00:00:00Z 欧盟：2026 年数字十年状况报告要求在 2030 年前弥合结构性差距 https://24-ai.news/zh/news/2026-06-18/github-copilot-app-ga/ 24 AI zh 2026-06-18T00:00:00Z GitHub：Copilot桌面应用正式发布——并行会话与云端自动化 https://24-ai.news/zh/news/2026-06-18/github-copilot-hydra-router/ 24 AI zh 2026-06-18T00:00:00Z GitHub：Copilot HyDRA路由器自动选择模型，节省72.5%成本且质量不变 https://24-ai.news/zh/news/2026-06-18/google-amie-primary-care-nature/ 24 AI zh 2026-06-18T00:00:00Z Google：AMIE在Nature研究中达到家庭医生水平的疾病管理能力 https://24-ai.news/zh/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI zh 2026-06-18T00:00:00Z IBM：研究显示91%的组织对AI依赖缺乏完整可见性，供应商锁定问题严峻 https://24-ai.news/zh/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI zh 2026-06-18T00:00:00Z OpenAI：基于GPT-5.4的近全自主AI化学家改进了药物合成中的关键反应 https://24-ai.news/ja/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI ja 2026-06-18T00:00:00Z Allen Institute：オープンソースのMolmoMotionが動画から3D動作を予測し、ロボティクスでSOTAを達成しました https://24-ai.news/ja/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI ja 2026-06-18T00:00:00Z AMD：オープンソース Schola が Unreal Engine と強化学習を接続し、ROCm 上でロボットアームをトレーニング https://24-ai.news/ja/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI ja 2026-06-18T00:00:00Z Anthropic：Claude Code v2.1.181 が /config 構文と Bun 1.4 を導入し、API 障害時の自動リトライを追加 https://24-ai.news/ja/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI ja 2026-06-18T00:00:00Z Anthropic：ソウルオフィスを開設し韓国AIエコシステムとの新たなパートナーシップを発表 https://24-ai.news/ja/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI ja 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens が示す法律 AI の幻覚率 52% という平均値に潜む 38 ポイントの格差 https://24-ai.news/ja/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI ja 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBenchが示すエージェントAIの疑似科学拡散——拒否率はほぼゼロ https://24-ai.news/ja/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI ja 2026-06-18T00:00:00Z AWS：Bedrock AgentCoreにウェブ検索、支払い機能、エージェントA/Bテストが追加されました https://24-ai.news/ja/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI ja 2026-06-18T00:00:00Z CNCF：KubernetesはエージェントAIの運用基盤——マルチエージェントセキュリティプラットフォームから得た教訓 https://24-ai.news/ja/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI ja 2026-06-18T00:00:00Z EU：2026 年デジタルの十年現状報告、2030 年目標に向けた構造的格差の是正を求める https://24-ai.news/ja/news/2026-06-18/github-copilot-app-ga/ 24 AI ja 2026-06-18T00:00:00Z GitHub：Copilotデスクトップアプリが一般提供開始——並行セッションとクラウド自動化 https://24-ai.news/ja/news/2026-06-18/github-copilot-hydra-router/ 24 AI ja 2026-06-18T00:00:00Z GitHub：Copilot HyDRAルーターがモデルを自動選択し、品質を維持しながら72.5%のコスト削減を実現しました https://24-ai.news/ja/news/2026-06-18/google-amie-primary-care-nature/ 24 AI ja 2026-06-18T00:00:00Z Google：AMIEがNature研究でかかりつけ医レベルの疾患管理能力を達成しました https://24-ai.news/ja/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI ja 2026-06-18T00:00:00Z IBM：調査で91%の組織がAI依存の完全な可視性を欠如——ベンダーロックインが深刻化 https://24-ai.news/ja/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI ja 2026-06-18T00:00:00Z OpenAI：GPT-5.4ベースのほぼ自律型AI化学者が薬物合成の重要反応を改善しました https://24-ai.news/ko/news/2026-06-18/allen-ai-molmomotion-robotics/ 24 AI ko 2026-06-18T00:00:00Z Allen Institute：오픈소스 MolmoMotion이 비디오에서 3D 움직임을 예측하고 로보틱스 분야에서 SOTA를 달성했습니다 https://24-ai.news/ko/news/2026-06-18/amd-schola-unreal-robotics-rl/ 24 AI ko 2026-06-18T00:00:00Z AMD：오픈소스 Schola가 Unreal Engine과 강화학습을 연결하여 ROCm에서 로봇 팔을 훈련합니다 https://24-ai.news/ko/news/2026-06-18/anthropic-claude-code-2-1-181/ 24 AI ko 2026-06-18T00:00:00Z Anthropic：Claude Code v2.1.181이 /config 문법과 Bun 1.4를 도입하고 API 연결 실패 시 자동 재시도 기능을 추가합니다 https://24-ai.news/ko/news/2026-06-18/anthropic-seoul-office-korea/ 24 AI ko 2026-06-18T00:00:00Z Anthropic：서울 사무소 개설 및 한국 AI 생태계와의 새로운 파트너십 발표 https://24-ai.news/ko/news/2026-06-18/arxiv-legalhallulens-legal-ai/ 24 AI ko 2026-06-18T00:00:00Z arXiv:2606.18021: LegalHalluLens가 법률 AI 환각률 52% 평균값 이면의 38 포인트 격차를 밝혀냅니다 https://24-ai.news/ko/news/2026-06-18/arxiv-pseudobench-agentic-pseudoscience/ 24 AI ko 2026-06-18T00:00:00Z arXiv:2606.18060: PseudoBench이 보여주는 에이전트 AI의 유사과학 확산——거부율 거의 제로 https://24-ai.news/ko/news/2026-06-18/aws-bedrock-agentcore-knowledge-update/ 24 AI ko 2026-06-18T00:00:00Z AWS：Bedrock AgentCore에 웹 검색, 결제 기능, 에이전트 A/B 테스트가 추가되었습니다 https://24-ai.news/ko/news/2026-06-18/cncf-cloud-native-agentic-security/ 24 AI ko 2026-06-18T00:00:00Z CNCF：Kubernetes를 에이전트 AI의 운영 기반으로——멀티 에이전트 보안 플랫폼의 교훈 https://24-ai.news/ko/news/2026-06-18/eu-digital-decade-report-2026/ 24 AI ko 2026-06-18T00:00:00Z EU：2026 디지털 10년 현황 보고서, 2030년까지 구조적 격차 해소를 촉구합니다 https://24-ai.news/ko/news/2026-06-18/github-copilot-app-ga/ 24 AI ko 2026-06-18T00:00:00Z GitHub：Copilot 데스크톱 앱 정식 출시——병렬 세션과 클라우드 자동화 https://24-ai.news/ko/news/2026-06-18/github-copilot-hydra-router/ 24 AI ko 2026-06-18T00:00:00Z GitHub：Copilot HyDRA 라우터가 모델을 자동 선택하여 품질 유지와 함께 72.5% 비용 절감을 달성했습니다 https://24-ai.news/ko/news/2026-06-18/google-amie-primary-care-nature/ 24 AI ko 2026-06-18T00:00:00Z Google：AMIE가 Nature 연구에서 가정의학과 의사 수준의 질병 관리 능력을 달성했습니다 https://24-ai.news/ko/news/2026-06-18/ibm-study-ai-dependency-lockin/ 24 AI ko 2026-06-18T00:00:00Z IBM：연구에서 91%의 조직이 AI 의존성에 대한 완전한 가시성 부재——벤더 잠금이 심각 https://24-ai.news/ko/news/2026-06-18/openai-ai-chemist-drug-synthesis/ 24 AI ko 2026-06-18T00:00:00Z OpenAI：GPT-5.4 기반의 거의 자율형 AI 화학자가 약물 합성의 핵심 반응을 개선했습니다