🌊 #HUST# alumnus Li Jun co-founded #DeepSea# #robotics# firm Sealien Robotics, which has raised over $70m in Series A funding. Its robots dive 1,000-6,000m and won repeat UAE orders after beating rivals from the UK, Netherlands, and Norway. #OVCTalent#

With Grok 3, we introduce two new features: DeepSearch and Think. DeepSearch is a powerful agent that can rapidly synthesize key information, reason about conflicting facts & opinions, and distill clarity from complexity

👑 Atria Dawn Preview is here, built to complete real research and engineering work. 📜 MIT. 🤖 ⚙️ Built on a 744B MoE foundation with a 256K context window. Standard and FP8 weights are available. 🔬 Discovery workflows cover evidence gathering, deep research, experiment design, execution, analysis, and recovery from failure. 🏆 Leads the reported comparison on AutomationBench, BFCL v4, CyberGym, DeepSearchQA, and BrowseComp. Scores include 53.8, 77.0, 86.5, 96.0, and 92.5 respectively. 🧩 Creation and delivery capabilities span software, interactive apps, ML systems, visualizations, reports, and presentations. 🛡 Cybersecurity support covers analysis, vulnerability validation, remediation, and retesting in authorized environments.

If you maintain a hand-built agent harness, this one is worth your time. (bookmark it) I feel like everyone is sleeping on the idea of dynamically generating agent harnesses on the fly. As you aim to own your harness, this is a topic more devs will lean into. Here is a great report discussing this topic. JIT-Agent is a model whose output is an agent harness. It formalizes the harness as a composable artifact under a fixed four-module protocol covering memory, planning, action protocol, and tool orchestration, then synthesizes one on the fly for any off-the-shelf agentic LLM. It also repairs harnesses mid-execution and self-evolves by distilling performance signals from an expanding archive of prior configurations. With JIT-Agent attached, DeepSeek-V4-Flash surpasses GPT-5.6 on DeepSearchQA (+9.1) and OdysseyBench (+4.3). GLM-5.2 gains up to +20.2 points. The generated harnesses are also performance-competitive with mature runtimes like OpenCode and Claude Code. Paper: Chat with Paper:

🌊 World's deepest floating wind platform connects to power grid The 16MW tension-leg floating wind platform "Haiyou Anlan" officially started supplying green power to offshore oil fields in the Pearl River Mouth Basin on August 6. It's the deepest and farthest of its kind globally, marking a world-class breakthrough for China's deep-sea wind tech.🌬️🌏 📍 Pearl River Mouth Basin, Guangdong #WindPower# #DeepSea# #CleanEnergy# #OffshoreWind# #HaiyouAnlan# #Guangdong# #Innovation# #RenewableEnergy# #ChinaTech# #FloatingWind# #GreenFuture#

🐟 "Gesheng No. 1": Offshore floating farm "Gesheng No. 1," a deep-sea aquaculture platform, was spotted off Xiaozhizhou Island, with the Hong Kong-Zhuhai-Macao Bridge visible in the distance. As part of Guangdong's push for smart ocean farming, this platform shows how technology is reimagining the way we grow our seafood. 🐟🌉 📍 Off Xiaozhizhou Island, Guishan Town, Zhuhai #DeepSeaFarming# #Aquaculture# #GeshengNo1# #Zhuhai# #OceanEconomy# #BlueFood# #SmartFarming# #Guangdong# #MarineTech# #Seafood# #Sustainability#

**From the black depths of an underwater cave, a colossal sea monster with elongated jaws and ghostly pale skin emerges into the light. Its terrifying presence freezes the diver who captured it. Is this a genuine giant of the deep, unseen for centuries, or the most realistic CGI ever filmed? The encounter has stunned the internet.** #DeepSeaMonster# #SeaCreature# #OceanHorror# #Cryptid# #UnidentifiedMarineLife# #AncientMysteries# #ViralTruth#

In 2023, scientists completed the most detailed 3D scan of the Titanic, using over 700,000 images to reconstruct the shipwreck. This digital model reveals new details about the ship's structure and the events of its final hours, offering fresh insights into the 1912 disaster. #Titanic# #MaritimeHistory# #DeepSeaExploration#

Agent performance isn't just about the model — the harness design matters just as much. TL;DR JIT-Agent dynamically synthesizes, repairs, and evolves agent harnesses (scaffolds) based on task characteristics at runtime. It achieves average gains of +7.7pt on GLM-5.2 and +8.8pt on DeepSeek-V4-Flash, reaching top-1 performance on 8 of 9 benchmarks — surpassing GPT-5.6 and all tested frontier models. Title: Scaling Harness Intelligence via Just-in-Time Harness Evolution URL: Key Points 🧩 Harnesses formalized as machine-learnable artifacts The four-module protocol h = (M, P, A, F) — Memory, Planning, Action, Capability Orchestration — constrains the generation space while remaining expressive enough to represent all 13 harnesses in HarnessFactory. 🎓 Three-stage training: imitation → repair → evolution Stage I learns from teacher-generated harnesses; Stage II trains recovery from execution failures (max 2 iterations); Stage III's Evo-GDPO evolves harnesses that advance the Pareto frontier on performance, latency, and cost simultaneously. 📊 Higher accuracy AND lower cost at the same time On xBench-DeepSearch: score 78→82 (+4pt), tokens 527K→212K (▲60%), cost $0.075→$0.039 (▲48%). Average 36% token reduction versus best fixed harness across all 9 benchmarks. ⚡ Transfers across model families without retraining JIT-generated harnesses outperform ReAct on DeepSeek V4 (+10.2pt avg), Mimo V2.5 (+8.6pt), and Qwen 3.6 (+4.0pt) — no need to retrain the harness generator for each backbone. 🔄 Online evolution continues improving at deployment Streaming mode accumulates successful harnesses across task sequences, outperforming static generation on all three evaluated benchmarks. "Harness intelligence" as a trainable scaling dimension orthogonal to model weights is the key conceptual contribution here. #AIAgents# #LLMScaling#