AI news
From Dataset Doubt to Instant Insight Every analyst knows the feeling: you open a dataset in your catalog, scroll through the columns, and wonder: “Is this even the data I need?” Until now, answering that meant writing exploratory SQL, checking lineage, or tracking down documentation. With the new Sample Data
AI news
Baidu World Conference 2024: A Shift in Focus and Philosophy From a personal demeanor standpoint, Robin Li seemed more relaxed than in previous years. From a business perspective, Baidu appeared to have intentionally narrowed its strategic focus. As a long‑time attendee of the Baidu World Conference, I immediately sensed
AI news
 --- # Datawhale Insights ## AI Agent Frameworks: Building Reliable Intelligent Applications To help learners start systematically, we’ve launched a series on AI agents: [**Hello-Agents Project Released — Learn Agents from Scratch**](https://mp.weixin.qq.com/s?__biz=MzIyNjM2MzQyNg=
AI news
VisionWeaver & VHBench-10 — Root Cause Diagnosis for LVLM Hallucinations Date: 2025-11-14 · Location: Shanghai The Bilibili User Technology Center has unveiled VisionWeaver and its diagnostic benchmark VHBench-10, offering a new paradigm for understanding and tackling hallucinations in large vision-language models (LVLMs). --- 📖 Preface For years, we’ve known LVLMs can misinterpret
AI news
Quantum Bit｜QbitAI Breaking the Bottleneck in Video Retrieval Current video retrieval research has reached a closed-loop bottleneck: For years, narrow-domain benchmarks like MSRVTT dominated, optimizing models for coarse-grained text queries. This led to: * Biased training data * Limited capabilities * Poor handling of fine-grained semantics * Weak long-context understanding * Inability to handle
AI news
Go 1.25: Introducing the Green Tea Garbage Collector Go 1.25 debuts a new experimental garbage collector, code-named Green Tea, delivering up to 40% faster performance for GC-intensive workloads compared to the current implementation. --- How Green Tea Works Green Tea retains Go’s mark-sweep strategy, but changes the
AI news
Improving Frontend Design with Claude Skills If you’ve ever asked a large language model (LLM) like Claude to quickly produce a web page — industry slang: “landing page” — chances are you’ve seen the same tired formula: * Inter font * White background with a purple gradient * Optional and forgettable animations Why
AI news
🎯 NeurIPS’25 Oral — InfinityStar vs. DiT A new NeurIPS’25 Oral paper from the ByteDance Commercialization Technology Team delivers a strong challenge to the Diffusion Transformer (DiT), which has long dominated video generation. --- Background — DiT’s Dominance and Drawbacks Ever since its debut, DiT has been the standard in
AI news
# November 13 — Beijing Humanoid Robot Innovation Center Fully Open-Sources Pelican-VL 1.0 ## Overview On **November 13**, the **Beijing Humanoid Robot Innovation Center** officially released **Pelican-VL 1.0**, a **fully open-source embodied Vision-Language Model (VLM)**. - **Parameter sizes**: 7B and 72B — *the largest open-source embodied multimodal model to date* - **Benchmark
AI news
Marble — The Viral 3D World Generator Is Now Open to All Yes — it's that 3D world generation model that went viral in the AI community just two months ago. Fei-Fei Li’s World Labs has now opened Marble to the public and released a full suite of powerful
AI news
# Doubao‑Seed‑Code: Breaking Free from Claude Code ## Overview Hello everyone, Over the past three months, I've felt that we’re getting closer to **fully breaking free from Claude Code**. ByteDance’s newly released **Doubao‑Seed‑Code model** has surpassed my expectations — not only in raw coding ability,
AI news
Produced | NetEase Technology OpenAI Shines Again: GPT‑5.1 Launches Just 3 Months After GPT‑5 Only three months after GPT‑5’s debut, OpenAI has announced GPT‑5.1 — positioning it as a major upgrade aimed at making ChatGPT “more intelligent and better at conversation.” Two brand‑new models