WebNews

Please enter a web search for web results.

NewsWeb

Mark Tech Post
marktechpost. com > 06/19/2026 > nvidia-ai-introduce-spatialclaw-a-training-free-agent-that-treats-code-as-the-action-interface-for-spatial-reasoning

NVIDIA AI Introduce Spatial Claw: A Training-Free Agent That Treats Code as the Action Interface for Spatial Reasoning

1+ hour, 7+ min ago  (483+ words) NVIDIA Research has released Spatial Claw, a training-free framework for spatial reasoning. It targets a persistent weakness in vision-language models (VLMs). These models still struggle to judge where objects are, how they relate, and how they move in 3 D. Spatial Claw…...

Symbols: nasdaq:nvda
Mark Tech Post
marktechpost. com > 06/19/2026 > vibethinker-3b-a-3b-dense-reasoning-model-built-on-qwen2-5-coder-3b-with-the-spectrum-to-signal-post-training-pipeline

Vibe Thinker-3 B: A 3 B Dense Reasoning Model Built on Qwen2. 5-Coder-3 B With the Spectrum-to-Signal Post-Training Pipeline

1+ hour, 52+ min ago  (749+ words) While recent breakthroughs in AI reasoning have largely been driven by massive scale, pouring in billions of parameters to cross complex cognitive thresholds'Vibe Thinker-3 B is charting a completely different path. Created by researchers from Sina Weibo Inc (China), this…...

Mark Tech Post
marktechpost. com > 06/19/2026 > liquid-ai-introduces-lfm2-5-embedding-350m-and-lfm2-5-colbert-350m-dense-bi-encoder-and-late-interaction-models-for-fast-multilingual-search-across-11-languages

Liquid AI Introduces LFM2. 5-Embedding-350 M and LFM2. 5-Col BERT-350 M: Dense Bi-Encoder and Late-Interaction Models for Fast Multilingual Search Across 11 Languages

13+ hour, 29+ min ago  (274+ words) Both target short-context search. Good fits include product catalogs, FAQ knowledge bases, and support docs. Liquid AI positions both as a drop-in replacement for an existing RAG pipeline. Both models start from LFM2. 5-350 M-Base, a mid-trained general-purpose checkpoint. Liquid AI applies…...

Symbols: nyse:estc,btc-usd
Mark Tech Post
marktechpost. com > 06/18/2026 > perplexity-launches-brain

Perplexity Launches Brain, a Self-Improving Memory System That Builds a Context Graph of an Agent's Work and Learns Overnight

1+ day, 3+ hour ago  (372+ words) Most AI memory remembers the user. It stores your preferences, your tastes, and your role. Perplexity is taking a different path. Today, Perplexity launched Brain, a self-improving memory system for its agent product, Computer. Brain does not focus on remembering…...

Symbols: btc-usd
Mark Tech Post
marktechpost. com > 06/18/2026 > the-kv-cache-compression-race-turboquant-vs-oscar-vs-epicache

The KV Cache Compression Race: Turbo Quant vs OSCAR vs Epi Cache

1+ day, 14+ hour ago  (440+ words) Long-context large language models (LLMs) face a memory bottleneck that has nothing to do with model weights. During decoding, transformers cache the key and value (KV) vectors for every token at every layer so they don't have to recompute attention....

Symbols: nasdaq:crwv
Mark Tech Post
marktechpost. com > 06/17/2026 > vercel-releases-eve

Vercel Releases Eve: An Open-Source AI Agent Framework Where Each Agent is a Directory of Files Mapped to Capabilities

2+ day, 6+ hour ago  (405+ words) Vercel has released eve, an open-source framework for building, running, and scaling agents. The project is published as the npm package eve, licensed under Apache-2. 0. Building an agent should mean defining what it does. It should not mean assembling all…...

Symbols: btc-usd
Mark Tech Post
marktechpost. com > 06/16/2026 > meet-qwen-robotsuite-three-embodied-ai-models-for-vla-manipulation-video-world-modeling-and-navigation

Meet Qwen-Robot Suite: Three Embodied AI Models for VLA Manipulation, Video World Modeling, and Navigation

3+ day, 7+ hour ago  (1335+ words) The Qwen team has released three embodied AI models, grouped as Qwen-Robot-Suite. The three are Qwen-Robot Manip, Qwen-Robot World, and Qwen-Robot Nav. Each is built on a Qwen vision-language backbone and targets a different robotics problem. Qwen-Robot Manip is a…...

Symbols: btc-usd
Mark Tech Post
marktechpost. com > 06/15/2026 > sakana-ai-marlin

Sakana AI Commercializes AB-MCTS in Sakana Marlin, an Enterprise Agent Generating Up to 100-Page Research Reports With Slides

4+ day, 1+ hour ago  (678+ words) Tokyo-based Sakana AI shipped its first commercial product "Sakana Marlin" this week. Sakana team positions it as a Virtual CSO (Chief Strategy Officer). It is a B2 B autonomous research agent built for enterprises. Marlin does not answer in seconds like…...

Symbols: nasdaq:rsss,btc-usd
Mark Tech Post
marktechpost. com > 06/15/2026 > meet-flash-kmeans-an-io-aware-exact-k-means-that-runs-over-200x-faster-than-faiss-on-gpus

Meet Flash-KMeans: An IO-Aware, Exact K-Means That Runs Over 200" Faster Than FAISS on GPUs

4+ day, 14+ hour ago  (712+ words) k-means has been an offline tool for decades. You run it once to preprocess data, then move on. A team of researchers from UC Berkeley and UT Austin released Flash-KMeans, a new open-source library that targets a different setting. Modern…...

Symbols: tpc-ds,ivf-pq,nasdaq:sitm
Mark Tech Post
marktechpost. com > 06/13/2026 > databricks-open-sources-omnigent-a-meta-harness-that-composes-governs-and-shares-ai-agents-across-claude-code-codex-and-pi

Databricks Open-Sources Omnigent: A Meta-Harness That Composes, Governs, and Shares AI Agents Across Claude Code, Codex, and Pi

5+ day, 18+ hour ago  (536+ words) Databricks released Omnigent, an open source "meta-harness" for AI agents. The project ships under the Apache 2. 0 license. The Databricks AI team built it with Neon. A harness is the wrapper around a model that turns it into an agent. Claude…...