Latest in AI and developer tools

Today's trail

updated 2m ago

Pulse · tbreak.com

tbreak.com · Nov 13

llama.cpp 1.0 reaches stable release for local AI

llama.cpp 1.0 is out. The local AI engine now signals breaking API and ABI changes with major-version bumps, and removes 35 deprecated functions.

Pulse · ntd.com

ntd.com · Nov 13

Nvidia CEO Resignation Sparks Chinese AI Shares Sell-Off and Renewed US Push to Curb Exports

The sell-off spilled over to major Chinese tech giants like Alibaba and Tencent, which rely on Nvidia chips to train their AI models. This week's rout erased a combined $75 billion in market value from the two companies, with their shares dropping 14 percent and 8 percent, respectively.

Pulse

Top picks this weekarXiv

arXiv

Nov 12

Back to Pixels: Lifting Latent Video Diffusion into Pixel Space

Frontier video generation models work in a VAE latent space that compresses video in both space and time. This compression made training tractable, but now limits high-resolution, controllable, and interactive generation -- doubling resolution quadruples token count, conditions apply to groups of frames rather than individual ones, and each response must wait for the rest of its multi-frame block. We present PixVid, a simple yet effective method that lifts a pretrained latent video model to directly generate every frame in pixel space. PixVid keeps the diffusion transformer and replaces the VAE decoder with a convolutional head that renders its outputs into pixels. We find that predicting clean video rather than velocity, together with a larger noise shift, is key to quality. Finetuned from Wan2.1-14B, PixVid is competitive with the latent baseline within $\sim 1.3$M training clips. Working framewise in pixels, it generates at $2\times$ higher resolution with the same per-frame token count, keeps depth-conditioned video editing aligned under fast motion, and responds to each action at the next frame. Project page: https://chenshuo20.github.io/PixVid/

Top picks

All arXiv Hugging Face GitHub Hacker News

Latest Top this week

arXiv

Sep 18

VideoReloc: Long-Term Indoor Video Relocalization against a Kilobyte-Scale Semantic Scene Graph

Given a compact semantic scene graph, long-term indoor video relocalization estimates a map-frame trajectory after lighting and furniture changes. Visual methods rely on appearance and become unreliable under these changes; localizing one frame at a time from object classes and…

arXiv

Sep 18

LLM-Generated Feature Pools for Time Series Anomaly Detection

AI LLM-generated features for time series anomaly detection

arXiv

Sep 18

A Principled Approach to Unsupervised Anomaly Detection

Traditional unsupervised anomaly detection (UAD) methods are designed to flag or localise deviations from a normative distribution, ignoring the underlying generative mechanisms of the anomalies. Yet the nature of an anomaly is often as important as its presence. We reformulate…

arXiv

Sep 18

An improved periodic activation for PINNs reconstructing convective flows

Architectures with periodic activation functions have already been shown to be beneficial in comparison to monotonic counterparts for a wide range of applications of physics-informed neural networks. Here, we investigate a network architecture which uses the complex exponential…

This week in Multimodal

This week

Get the morning digest

Today's trail

Pulse

Top picks

Today's trail

Rising tools

This week in Multimodal

Most active this week