Official account of Tencent Technology. A platform for publishing and analyzing Tencent's technological innovations and cutting-edge developments.
SpecExit combines speculative sampling with a lightweight draft model to predict early‑exit signals, shortening large‑reasoning model chains by up to two‑thirds and achieving up to 2.5× end‑to‑end inference acceleration on vLLM without sacrificing accuracy.
This article explores the strategic impact of AI‑assisted coding, outlines a comprehensive AI × SDLC methodology, showcases three real‑world scenarios—from end‑to‑end development pipelines to Figma‑to‑code automation and backend system evolution—while highlighting practical tips, pitfalls, and organizational changes needed to fully harness AI programming.
This comprehensive guide introduces Rust’s core syntax and unique programming mindset, walks through a practical greplite example, and covers essential topics such as installation, Cargo commands, data structures, ownership, traits, iterators, async/await, concurrency, and best learning resources for rapid mastery.
Tequila introduces a novel 1.58‑bit ternary quantization for large language models that tackles the dead‑zone trap by reactivating zero‑weight biases with dynamic offline offsets, achieving near‑full‑precision performance, faster convergence, and up to three‑fold CPU inference speedups.
This article reveals ten common C++ performance pitfalls—from costly abstractions and virtual functions to hidden copies, destructors, and misuse of smart pointers—while also showcasing practical language features and optimization techniques that can dramatically improve runtime efficiency in production code.
An in‑depth investigation reveals the rise and takedown of the RapperBot DDoS botnet, detailing its malware lineage, sample analysis, sophisticated attack techniques, criminal profit models, and practical security recommendations, while showcasing Tencent’s Zeus Shield intelligence platform and AI‑enhanced threat analysis.
Discover practical insights from a front‑end maintainer on extending the lifespan of open‑source projects, covering community engagement, decisive leadership, progressive migration, trade‑off handling, documentation, contributor recognition, and product‑like management to prevent project death and foster sustainable growth.
The POINTS-Reader paper, accepted at EMNLP 2025, introduces a two‑stage, fully automated data generation pipeline that enables a lightweight visual‑language model to extract text, tables, and LaTeX formulas from diverse PDF layouts with superior performance and high throughput, all without relying on costly teacher‑model distillation.
Learn how to craft effective prompts for large language models by using a systematic structure—role and task, core principles, context handling, chain‑of‑thought, output specifications, and few‑shot examples—and discover techniques for generating and iteratively refining prompts with the model itself.
The article outlines ARC Lab’s evolution from its 2019 founding as an internal corporate research unit to a high‑impact AI team that pursues difficult multimodal understanding and generation problems, measures success through a technology‑impact funnel, publishes 30‑40 top‑tier papers annually, and translates research into open‑source tools and products that drive academic, industry, business, and societal value.
