All Updates
Page 125 of 1439
June 17, 2026
New Architecture for Decentralized Multi-Agent Collaboration Networks
This research proposes a layered architecture for peer-to-peer agent networks, enabling heterogeneous agents to discover, trust, and cooperate on complex tasks. It addresses critical challenges like semantic announcement propagation and verifiable identity governance.
MemTrace: New Benchmark for LLM Long-Term Memory Accuracy
MemTrace is a new benchmark that evaluates LLM long-term memory by focusing on 'knowledge points' rather than individual questions. It reveals that memory failures are primarily due to poor evidence utilization rather than retrieval limitations.
DivInit: Boosting Agentic Search via Diverse Query Initialization
DivInit is a training-free intervention that improves agentic search by diversifying initial queries. By selecting diverse seeds from a single model call, it reduces retrieval redundancy and improves performance on multi-hop QA tasks.
China Regulator Encourages AI Firm IPOs to Boost Markets
China's securities regulator is actively encouraging artificial intelligence developers and Hong Kong-listed companies to pursue domestic IPOs. This initiative aims to strengthen onshore capital markets by integrating high-growth technology firms.
Microsoft Challenges Multi-Vendor Email Security Strategy
Microsoft released benchmarking data suggesting that integrated email security partners provide only marginal improvements over its native Defender for Office 365 protections. While Microsoft claims superior threat detection, industry experts argue that a defense-in-depth approach remains necessary for enterprise security.
ASUS Announces ROG 20th Anniversary Limited Edition Hardware
ASUS is launching a limited edition ROG 20th anniversary collection featuring high-end hardware including the RTX 5090 D v2 GPU and AMD Ryzen 9 9950X3D2. The flagship bundle is priced at 112,026 RMB and will be available via a lottery-based offline event in Shanghai.
Lingyi iTech Targets $1.1B Hong Kong IPO
Chinese manufacturer Lingyi iTech Guangdong Co. is initiating an IPO in Hong Kong. The company aims to raise up to $1.1 billion to support its operations.
Microsoft Teams Efficiency Mode Optimizes Performance for 8GB RAM
Microsoft is rolling out an 'Efficiency Mode' for Teams to improve performance on resource-constrained hardware. By optimizing video resolution and startup processes, the update significantly reduces memory usage for devices with 8GB of RAM.
Intel unveils three pillars for future chip innovation
Intel has detailed its future roadmap focusing on three key technologies: CFET, Gallium Nitride (GaN) on silicon, and Ruthenium interconnects. These innovations aim to push the boundaries of process node scaling.
Molecular Heart secures over $100M for AI protein research
Led by Xu Jinbo, Molecular Heart has raised over $100 million to build infrastructure for the global AI protein industry, aiming to lead the second paradigm revolution in AI biology.
GCash Owner Receives Approval for Philippine IPO
The parent company of the Philippines' leading fintech app, GCash, has received regulatory approval for an IPO. Globe Telecom Inc. confirmed the milestone for the digital wallet provider.
Korean Investors Snap Up $800M in SpaceX Shares
South Korean retail investors purchased $800 million worth of SpaceX shares on the first day of trading. This reflects high demand for the company despite limited access to traditional IPOs.
General Atlantic Leads Funding for Chinaโs Kling AI
Kuaishou Technology is in talks with General Atlantic to lead a funding round for its video AI unit, Kling AI. The move aims to secure a US backer before a potential IPO.
HSBC Partners with Google Cloud to Scale AI Operations
HSBC Holdings Plc has announced a strategic partnership with Google Cloud to integrate artificial intelligence across its global banking operations. The bank expects each implemented AI project to generate over $100 million in revenue or efficiency gains.
WeChat Pay to Launch AI-Exclusive Card This Week
WeChat Pay is reportedly launching an 'AI-exclusive card' as early as this week. This move signals a deeper integration of AI-driven financial services within the WeChat ecosystem.
Weibo's 3B Model Challenges AI Scaling Laws on Benchmarks
Sina Weibo researchers introduced VibeThinker-3B, a 3-billion parameter model that allegedly matches or exceeds the reasoning performance of massive flagship models. The release has sparked intense debate within the AI community regarding the validity of current benchmarks.
ByteDance Launches Seedance 2.0 Mini Model
ByteDance has officially released Seedance 2.0 Mini, a new iteration of their generative AI model. This launch highlights the company's ongoing expansion in the competitive AI model landscape.
OpenAI Launches LifeSciBench for AI Life Science Evaluation
OpenAI has introduced LifeSciBench, a new benchmark specifically designed to evaluate how AI models perform on complex, real-world life science research tasks. It features expert-authored and reviewed content to ensure high-quality assessment of scientific decision-making.
Murata releases simulation models for Ansys HFSS and Icepak
Murata Manufacturing has launched a new service providing simulation models directly compatible with Ansys HFSS and Icepak tools. Users can now access and download these models via the Murata website to streamline electromagnetic and thermal analysis workflows.