chinese

Tag

Cards List
#chinese

Beyond Atomic Tokens: Factorizing Syllables for Language Model Pretraining

arXiv cs.CL ↗ · 6d ago Cached

This paper introduces a linguistically motivated phonemic tokenizer for Vietnamese and Chinese that factorizes syllables into onset, rime, and tone components, reducing vocabulary size and improving efficiency. The proposed PhonemicBERT models demonstrate competitive or superior performance in language understanding tasks compared to existing tokenizers and pretrained models.

0 favorites 0 likes
#chinese

HSK Manga: Turning Manga into Beginner-Level Chinese

Lobsters Hottest ↗ · 2026-08-30 Cached

HSK Manga is a free tool that simplifies manga dialogue to HSK levels for Chinese learners, using AI and OCR to make reading more accessible for beginners.

0 favorites 0 likes
#chinese

@crblandet: Speaking of the ICT trading model, a while ago I compiled a Chinese tutorial using Fable 5 (burned through most of my weekly quota) and put it on my website. First author is Fable 5, corresponding author is me. Welcome to human or AI distillation https://chanlunpp.org/basics-ict

X AI KOLs Timeline ↗ · 2026-08-14 Cached

The Chinese tutorial on the ICT trading model, created using Fable 5, is now online, systematically explaining the core concepts and trading system.

0 favorites 0 likes
#chinese

@XAMTO_AI: Want to write a coding agent yourself? You've read plenty of articles about the principles, but you get stuck at the first line when you actually start. Recently I found the Chinese tutorial *Hands-On Pi*, which starts with a real offline agent log and takes you through all 15 levels from the beginning. Truly build a complete coding agent from scratch—every step has runnable code...

X AI KOLs Timeline ↗ · 2026-08-13 Cached

Introduces the Chinese tutorial *Hands-On Pi*, covering the step-by-step implementation of a Pi-style coding agent from scratch through 15 runnable checkpoints, including protocols, tool loops, conversation tree storage, context compression, and evaluation. Each chapter comes with the main text, a corresponding commit, tests, and failure experiments.

0 favorites 0 likes
#chinese

@Suu766: You've already used a VPN to get on Telegram, but still can't search for resources — it's really a waste!!! Many people are looking for groups, channels, and bots, and still begging in the comments for someone to send links. Actually, someone has already organized Chinese TG resources for you. Channels, groups, bots, tool resources — you can find them directly by searching keywords, saving you from searching high and low in Telegram…

X AI KOLs Timeline ↗ · 2026-08-09 Cached

Recommend using tgzhidao.com, a Chinese Telegram navigation site. Users can search channels, groups, bots, and tool resources by keywords, avoiding the need to search for a needle in a haystack in Telegram.

0 favorites 0 likes
#chinese

@grgerwcwetwet: Played Telegram for a long time, still don't know where to find quality channels? Recommend a practical website TG知道, which specifically collects Chinese Telegram channels, groups, and bots. It covers tech & digital, learning & tech, news, communities, practical bots, etc. Direct category search, much more convenient than randomly searching around in Telegram yourself…

X AI KOLs Timeline ↗ · 2026-08-08 Cached

Recommend a website called TG知道, which specifically collects Chinese Telegram channels, groups, and bots, supports categorized search, making it convenient for users to discover quality resources.

0 favorites 0 likes
#chinese

Hy-MultiTurn: A Six-Dimensional Benchmark for Deep Multi-Turn Dialogue Understanding

arXiv cs.CL ↗ · 2026-08-03 Cached

Introduces Hy-MultiTurn, a Chinese benchmark for deep multi-turn dialogue understanding, evaluating 22 frontier models across six controlled modes (constraint memory, precise execution, constraint synthesis, object localization, action suppression, reference resolution) with 209 tasks spanning 12-76 turns. Even the strongest model, GPT-5.5, satisfies only 41.1% of responses.

0 favorites 0 likes
#chinese

@jinchenma_ai: Removing AI flavor is not something that can be solved with a set of prompts or an open-source skill. If you want to remove AI flavor as cleanly as possible and make your writing more like yourself, you need to build an iterative, compounding writing system for yourself.

X AI KOLs Timeline ↗ · 2026-08-02 Cached

Jinchenma shares how to remove AI flavor through an iterative personal writing system, including clarifying author identity, article types, taboo rules, and methods for multiple independent reviews and rule iteration.

0 favorites 0 likes
#chinese

@seclink: mimofan has built-in the following 29 features: 1. Tool call loop breaker (DEFAULT_MAX_CONTINUATIONS = 50) 2. Auto-compaction mechanism (http://compaction.rs) 3. Token usage tracking (T…

X AI KOLs Timeline ↗ · 2026-07-29 Cached

The mimofan tool has built-in 29 features, including tool call loop breaker, auto-compaction mechanism, token usage tracking, system prompt optimization, etc. It is a feature-rich AI agent tool.

0 favorites 0 likes
#chinese

CallBench: A Benchmark for Dual-Goal Coordination in Phone Call Assistants

arXiv cs.AI ↗ · 2026-07-28 Cached

CallBench is a Chinese benchmark for evaluating dual-goal coordination in phone call assistants, containing 50,000 multi-turn dialogues across six scenarios, with a preset-aware evaluation protocol covering semantic understanding, safety, and dialogue rhythm.

0 favorites 0 likes
#chinese

@NFTCPS: Just found a treasure on GitHub called TelegramGroup. It neatly categorizes the scattered Telegram groups, channels, and bots. The biggest headache when I first started using Telegram was finding nothing—either ads or dead groups. This project lines them up by category: Tech, Software, AI...

X AI KOLs Timeline ↗ · 2026-07-21 Cached

Introduces a GitHub project called TelegramGroup that curates a large collection of high-quality Telegram groups, channels, and bots, categorized by direction to help newcomers quickly find the resources they need.

0 favorites 0 likes
#chinese

@Fenng: 腾讯混元 Hy3 这次站稳了

X AI KOLs Following ↗ · 2026-07-15 Cached

腾讯混元 Hy3 模型被认为这次表现稳定,有显著进步。

0 favorites 0 likes
#chinese

@QingQ77: Want to systematically learn AI Agent but afraid of giving up halfway? This Chinese tutorial website turns 21 chapters into a visualized textbook with animated flowcharts. Each chapter comes with a quiz bank and an AI teaching assistant ready to answer questions anytime. The entire site is static and can be deployed to Nginx with a single command. Content is spread across 6 major sections covering 21 chapters, starting from LLM basics, Func...

X AI KOLs Timeline ↗ · 2026-07-14 Cached

Recommends an open-source Chinese AI Agent general knowledge tutorial website, including 21 chapters with visualized animated tutorials, a 130+ question bank, and an AI teaching assistant. It is purely static and can be deployed to Nginx with one click.

0 favorites 0 likes
#chinese

MedRealMM: A Real-World Multimodal Benchmark for Chinese Online Medical Consultation

arXiv cs.AI ↗ · 2026-07-13 Cached

MedRealMM is a new multimodal benchmark for Chinese online medical consultation, built from real-world patient-doctor interactions, evaluating LLMs on next-response generation with clinical rubrics.

0 favorites 0 likes
#chinese

@RealCodedAlpha: You might have been prompting GPT-5.6 the old way. OpenAI has actually released the latest official guide. In this guide, OpenAI clearly explains task descriptions, constraints, tool calls, output formats, and common pitfalls. I read the entire official documentation and compiled it into an accessible Chinese version…

X AI KOLs Timeline ↗ · 2026-07-12 Cached

A curated Chinese summary of OpenAI's official GPT-5.6 prompt guide, including core changes, recommended structure, templates, and common pitfalls.

0 favorites 0 likes
#chinese

@wlzh: On the same day, I uncovered two gems — Liu Run's Business Reference Distillation Skills and Luo Zhenyu's Quote Distillation Skills. What does that mean? It's not about learning how Liu Run and Luo Zhenyu write — it's about turning their writing systems into templates that AI can use directly. That is, you feed in a piece of material, and AI will use Liu Run's methods to deconstruct the structure, create analogies, and build SCQA logic chains; then switch to Luo Zhenyu's mode to generate quotes, insights, and personal engagement.

X AI KOLs Timeline ↗ · 2026-07-11 Cached

Introduces two open-source AI writing skill templates that mimic Liu Run's business analysis style and Luo Zhenyu's quote style, enabling AI to produce professional and shareable articles, lowering the barrier to writing.

0 favorites 0 likes
#chinese

@xingbugengming: I don't have the GPT 5.6-Sol model, but I spent an hour using up five hours' worth of quota to test the frontend capabilities of GPT 5.6-Terra. The results were very impressive, especially when handling image matting—it's amazing. I asked it to create a webpage: an interactive card of Guan Yu from the Three Kingdoms, with mouse drag...

X AI KOLs Timeline ↗ · 2026-07-10 Cached

The user tested the frontend capabilities of GPT 5.6-Terra, creating an interactive Guan Yu card webpage from the Three Kingdoms. The results were outstanding, particularly in image matting and depth-of-field perspective.

0 favorites 0 likes
#chinese

@cyrilxuq: I've released a free open-source "Investment Guide for Beginners". Interested folks can star it. Our ops team has organized everything yesterday, and we'll keep adding new technical terms. Today I'll also release a major update with "Investment Lesson 1", covering the few paths ordinary people can take to succeed in investing.

X AI KOLs Timeline ↗ · 2026-07-08 Cached

Xu Chonglang released a free open-source "Investment Guide for Beginners" covering basics of US stocks, options, and cryptocurrency, aiming to help Chinese investors build a complete knowledge framework, identify risks, and make independent decisions.

0 favorites 0 likes
#chinese

@passluo: Let me brag a bit. After CC Max delists fable 5 tonight, we'll most likely continue to offer it at the same price for now. Tested today, error rate is about 3%. Let's see if it blows up tonight and proves me wrong.

X AI KOLs Timeline ↗ · 2026-07-07 Cached

CC Max will continue to offer the service after delisting fable 5, at the same price. Current test error rate is about 3%, but concerned about potential issues after launch.

0 favorites 0 likes
#chinese

@gkxspace: Chinese people are damn amazing — one website vs. all social platforms fulfills all my data fantasies. Indeed, where there is demand, there is a market.

X AI KOLs Timeline ↗ · 2026-07-03 Cached

余温 shares a review of his three-year AI startup journey, arguing that technology is not valuable; channel and reach capabilities are what truly matter. He also recommends a website that aggregates data from all social platforms.

0 favorites 0 likes
Next →
← Back to home

Submit Feedback