Dikemas kini setiap hari3937 cerita yang disahkan
Berita AI. Tanpa bunyi bising.
Liputan AI yang diperiksa sumber mengenai pelancaran produk, perubahan dasar, penyelidikan keselamatan, dan langkah industri, dijelaskan dalam bahasa Inggeris mudah oleh pasukan pendidikan bukan untung.
Sumber yang disahkan
Setiap cerita berhubung dengan bukti terkuat yang ada: sumber asal apabila ada, laporan yang sebaliknya jelas dikaitkan.
Bahasa Inggeris biasa
Apa yang berlaku, mengapa ia penting, dan perkara yang perlu ditonton — tanpa jargon.
Tiada pengisi
Apabila isyarat nipis, kami tidak menerbitkan apa-apa selain daripada mengalas suapan.
Lagi cerita
9 ceritaproduk
Goodfire presents Silico, an agent for interpreting and debugging AI models
Goodfire’s new Silico platform is designed to inspect hidden model representations, diagnose behavior and guide targeted interventions across life-sciences, robotics, vision and language models.goodfire.comInovasi
CoDrift proposes compositional action fields for offline reinforcement learning
A new arXiv preprint introduces CoDrift, a one-step generative policy framework that combines behavioral compatibility and value-seeking objectives for offline reinforcement learning. The authors report the best average rank across 73 OGBench and D4RL tasks in both offline and offline-to-online settings.arxiv.orgInovasi
Preprint proposes a Bayesian way for LLM agents to recognize when they need stronger help
A new preprint studies agents that can transfer control to a stronger language model during reasoning, combining a Bayesian stopping rule with theoretical guarantees and a limited Qwen2.5-Coder validation.arxiv.orgproduk
Ars Technica melaporkan IBM melancarkan model berat terbuka Granite 4.2 untuk AI tempatan
Keluarga Granite 4.2 IBM termasuk model berat terbuka 3B, 8B dan 30B dengan tetingkap konteks 128,000 token asli. Ars Technica melaporkan bahawa dua versi yang lebih besar menerima latihan khusus untuk kegunaan alat dan tugas agen.arstechnica.comInovasi
PinSieve reports production gains from selective VLM serving and governed memory
A KDD 2026 workshop paper describes a deployed vision-language-model serving agent that handles unresolved content-quality cases, reporting higher review productivity, lower normalized operating cost and faster signal delivery.arxiv.orgPerusahaan
loveholidays berkata Codex meletakkan pembinaan perisian dalam jangkauan bukan jurutera
Loveholidays syarikat pelancongan dalam talian berkata Codex membantu pasukan bukan kejuruteraan membina pengalaman pelanggan dan perubahan infrastruktur sambil meningkatkan perubahan dan penggunaan kod dibantu AI.
openai.comInovasi
PuzzleKV proposes page-wise compression to reduce LLM KV-cache storage
A new arXiv paper proposes PuzzleKV, a training- and calibration-free method that independently compresses completed pages of large language model KV caches. It reports more than 96% of Full KV performance at about 60% of original storage, and more than 93% at 18.7% of original storage with quantization.arxiv.orgInovasi
A bandit framework cuts LLM calls for automated essay-scoring prompt selection
A new arXiv paper reports that a multi-armed bandit controller selected prompting strategies for LLM essay scoring with comparable accuracy to exhaustive search while reducing calls by 78.4%.arxiv.orgDasar
Bill Gates says AI has crossed multiple danger thresholds, MIT Technology Review reports
In an MIT Technology Review interview, Bill Gates says frontier AI has crossed thresholds involving biological misuse, cyberattacks, psychosocial dependence, job displacement and loss of control. He proposes monitoring models capable of creating novel molecules, international cooperation and new approaches to…technologyreview.com
Satu taklimat berguna setiap minggu
Teruskan AI tanpa perlu hidup dalam suapan itu.
Dapatkan berita AI yang disahkan minggu ini, data asal, alat berguna, pilihan pembelajaran, dan pekerjaan AI segar.
Jangkau orang yang belajar AI
Mengupah profesional AI atau melancarkan produk AI yang berguna? Tunjukkan kepada orang yang datang ke sini untuk belajar dan bertindak.
Siarkan kerja AIHantar alat AI