#Research
Curated coverage on «Research»: releases, research, benchmarks, open-source.
Deli's evolution: how a stationery manufacturer is capturing the high-tech market
China's Deli is breaking stereotypes about "low-tech" stationery. From developing unique pen tips to building its own printers based on domestic chips.
TI-DPO: A New Method for AI Alignment by Evaluating Token Importance
Researchers introduced TI-DPO, an algorithm that fixes the main problem of standard DPO through detailed analysis of each word during neural network training.
Ant Group open-sources multimodal model Ming-Flash-Omni 2.0
Ant Group unveiled Ming-Flash-Omni 2.0. The new open-source model challenges Gemini 2.5 Pro, offering advanced text, image, and voice capabilities in a single…
Shanghai scientists uncover the "dark side" of social interactions among AI agents
Diffusion language models challenge GPT with a 892-token speed record
First the brain, then the dumplings: Alibaba DAMO Academy improves robot intelligence
Goodbye, latency: Nanyang Technological University speeds up robot response in dynamic environments
Self-Distillation: How Self-Distillation Will Change AI in 2026
# Self-Distillation: How Self-Distillation Will Change AI in 2026 The artificial intelligence industry has reached a critical juncture.
Lingxi Photonics raised millions to develop optical engines for AI clusters
# Optical Engines Instead of Copper Cables: How a Chinese Startup Solves a Critical Problem for AI Data Centers When you turn on a Netflix movie, video content is…
ProjDevBench: Can AI Build Full-Fledged Software from Scratch?
SGLang Ecosystem: Outcomes of a Large-Scale Developer Meeting in Shanghai
Neural network trained to check and rewrite code
New RRAM Memory Breaks Through AI Performance Limitations
Kimi K2.5 leads in OpenRouter rankings, outpacing Gemini and Claude
The Chinese artificial intelligence market continues to impress. The latest breakthrough is the Kimi K2.5 model, developed in China and demonstrating impressive…
Call for Submissions for CVPR 2026 Workshop: Security of Multimodal Agents
The artificial intelligence world is developing rapidly, and with it grows the need to ensure the safety and reliability of the systems being created.