温馨提示：本站仅提供公开网络链接索引服务，不存储、不篡改任何第三方内容，所有内容版权归原作者所有
AI智能索引来源：http://www.bee.com/ar/51895.html
点击访问原文链接

Yupp raises $33 million in seed round, former Twitter technical backbone reconstructs AI model evaluation system | Bee Network

进入原网站 导航首页
Yupp raises $33 million in seed round, former Twitter technical backbone reconstructs AI model evaluation system | Bee Network Login الأخبار الشائعة منصة إطلاق ميمي وكلاء الذكاء الاصطناعي ديسي مستكشف السلسلة الأعلى لنوبي 100x عملات معدنية لعبة النحل المواقع الأساسية يجب أن يكون لديك التطبيق مشاهير التشفير ديبين الناشئين الأساسية كاشف الفخ الأدوات الأساسية المواقع المتقدمة التبادلات أدوات NFT أهلاً، خروج عالم الويب 3 ألعاب تطبيق خلية نحل منصة النمو إعلان يبحث إنجليزي إعادة شحن العملات تسجيل الدخول تحميل ويب 3 يوني ألعاب تطبيق خلية نحل إعلان بيت•تحليل•Yupp raises $33 million in seed round, former Twitter technical backbone reconstructs AI model evaluation system Yupp raises $33 million in seed round, former Twitter technical backbone reconstructs AI model evaluation systemتحليلمنذ 9 أشهرجديدوايت 39٬239 7 Original author: KarenZ, Foresight News

With the rapid development of AI technology, the emergence and diversification of AI models dazzles users. How to choose an AI model that suits you, how to make the model accurately capture real needs in addition to traditional benchmark tests, and how to provide real incentives for feedback users have become key issues that the AI industry needs to break through.

Yupp is an open platform born in this context, aiming to build an open, transparent, community-driven AI model evaluation platform. As Yupp said, More than any other technological innovation in history, AI relies on everyones participation and contribution to drive evolution.

Last week (June 13), Yupp.ai announced a $33 million seed round led by a16z تشفير, and a slew of investors including Google Chief Scientist Jeff Dean and X co-founder Biz Stone attracted the attention of the community. At the same time, Yupp also launched its products, which not only provide users with a window to explore AI, but also reتحديne the evaluation and optimization of AI models through community participation and blockchain technology.

Yupp team and financing background The company behind Yupp is Ber Sarai Labs Inc., which was co-founded by Pankaj Gupta and Gilad Mishne in June 2024 and has been testing it secretly for the past six months. The two co-founders and chief scientists of Yupp met at Twitter in 2010. They both have a deep background in the AI industry and have worked at companies such as Coinbase, Google, and X.

Pankaj Gupta: Co-founder and CEO of Yupp, holds a bachelors degree in computer science and engineering from the Indian Institute of Technology, Delhi, and a PhD in computer science from Stanford University. He has served as technical director, senior manager of personalization and recommendations, and senior ML staff at Twitter (March 2009-May 2014), engineering director and senior engineering director at Google (July 2017-March 2021), and served as the number one employee and site head of Coinbase India, and later served as vice president of engineering and consultant at Coinbase (April 2021-May 2024).

Gilad Mishne: Yupp Co-founder and Head of AI, previously worked as a software engineer at Intel (1998-2000), a senior scientist at Yahoo (2017-2010), a senior engineer and director of search at Twitter (2010-2015), and a senior engineering manager and head of machine learning at Googles Moon Factory at Google (2019-2023).

Jimmy Lin: Chief Scientist at Yupp. He studied question-answering systems and conversational interfaces while pursuing his PhD in computer science at MIT. He is currently a professor and David R. Cheriton Chair at the David R. Cheriton School of Computer Science at the University of Waterloo. From 2010 to 2012, Jimmy Lin was involved in data analysis and data science infrastructure construction at Twitter.

The 33 million seed round of financing announced by Yupp this month was completed last year. Yupps capital matrix covers technology, investment and academia. In addition to the lead investor a16z crypto, Google Chief Scientist Jeff Dean, X co-founder Biz Stone, Pinterest co-founder Evan Sharp, Perplexity CEO Aravind Srinivas, Cred CEO Kunal Shah, four professors from Stanford University (Dan Boneh, Chris Re, Nick McKeown, Balaji Prabhakar), Othman Laraki, Paul Grewal, Gokul Rajaram and Coinbase Ventures also participated.

What is Yupp and how does it work? Yupp is positioned as an AI model exploration and evaluation platform, allowing users to experience and compare various AI models for free. Its core concept is to evaluate models through crowdsourcing: users submit prompts, compare the responses generated by different AI models, and give better answers and evaluation feedback, while receiving redeemable points incentives. These choices and feedback will be recorded to form data for later training and evaluation of AI models.

Yupp will also leverage open access and permissionless technologies such as blockchain, cryptographic primitives and protocols such as zero-knowledge proofs and challenge/response mechanisms, and privacy-preserving technologies such as confidential computing to build systems with provable trustworthy neutrality, fairness, and robustness.

The operating mechanism of Yupp can be summarized as follows:

1. Model exploration and comparison: Yupp has collected more than 500 AI models, including ChatGPT, Claude, Gemini, DeepSeek, Grok, Llama, etc. Users can find various AI models on the platform chat page and conduct prompt tests to intuitively compare their pros and cons side by side.

Yupp currently has two pages: the chat page and the ranking page. The chat page is relatively simple in design, with a message box, file upload function, model selection function (optional), image upload function, and chat private and public mode switching (private by default).

As shown in the figure below, after asking the question, Yupp will give two AI answer versions, and then the user can choose the better version.

It is worth mentioning that the model is automatically selected by default when asking questions, and sometimes the model name is hidden to collect more objective feedback. Of course, users can also ask questions randomly. In addition, Yupps QuickTake AI function can also provide a brief summary response.

2. User feedback: After selecting the best answer, users can further provide feedback on the quality of the answer by clicking on the label, or they can freely express their preferences in words. This feedback will help users customize the AI answers on Yupp in the future and also enable Yupp to provide models for free.

3. Feedback Rewards: After giving feedback, users will get a points scratch card. Points can be used to ask questions or cash out.

4. Evaluation: Users participate in the model evaluation process by selecting the best response and providing feedback. Yupp aims to establish an open and transparent evaluation system so that AI developers can obtain valuable training data, users can get rewards, and jointly promote the development of AI technology. The Yupp platform has a public leaderboard called Yupp VIBE Score (VIBE, Vibe Intelligence BEnchmark), which aims to use user feedback to improve model performance while ensuring the privacy of user prompts unless users choose to share.

Yupp will create a ranking list based on user feedback and response speed. In the ranking list, the AI models integrated on the platform can be sorted by filters, including VIBE score, confidence interval of probability sample, voting status, speed, latency, and input/output cost.

Yupps AI model evaluation combines user preference data to segment user and evaluation data in a more granular manner, thereby providing samples for AI developers. Yupp said that with the team founders experience in dealing with spam and robot problems on Twitter, a complex algorithm was developed to eliminate low-quality data and ensure the integrity of the ranking. Yupp has also formed a dedicated trust and safety team and will continue to invest heavily in this area.

Yupp Points System: Balanced Rules of Consumption and Redemption Yupp points are consumed by asking questions, earned by giving feedback, and partially redeemable. Yupp said that if users use Yupp responsibly, they will always have enough points to ask questions to the AI model, and can also redeem a portion of the points to show their gratitude for improving the ecosystem.

Asking questions costs points, and you can get 5000 points for free after initial registration. In addition, the total fee is the sum of the default fee, PRO model fee, attachment submission fee, and pre-selected image model fee.

Basic Fee: Each clue word costs 50 points by default. If you generate an image, the fee is 100 Yupp points.

PRO Models: These are advanced models, available only on other platforms through paid subscription. If the user pre-selects the PRO model, it will cost an additional 50 Yupp points per question.

MAX Models: These models are the most expensive models to use. If the user pre-selects the MAX model, then each model will cost an additional 300 Yupp points per prompt word (total 350 points including the base cost).

Attachment Submission Fee: Each attachment costs 25 Yupp Points.

Selecting an Image Model: Each pre-selected image model costs an additional 100 Yupp Points.

In addition, Yupp QA is private by default. If users choose to make the QA public, they only need to pay half of the regular fee mentioned above.

As mentioned earlier, feedback models can earn points scratch cards. The few points scratch cards I got ranged from 200 to 500 points.

Yupp said that users can withdraw points and exchange them for US dollars, euros, Indian rupees and more than 20 other currencies, or exchange them for stablecoins (based on Base and Solana). Yupp has also established partnerships with payment service providers such as Stripe, Paypal and Coinbase to meet the different needs of users. Every 1,000 points can be exchanged for 1 US dollar. However, the redemption function is currently not available. In order to avoid witches or abuse, Yupp has also set rules for redeeming points:

The maximum number of withdrawals per day is 1, and the maximum withdrawal amount is US$10 (10,000 points);

The maximum number of withdrawals per week is 3, and the total withdrawal limit is US$20 (20,000 points);

The maximum number of withdrawals per month is 6, and the total withdrawal limit is US$50 (50,000 points).

In addition, Yupp stipulates that any purchase, sale, trade or transfer of Yupp Points, etc., which violates the Terms of Service, will be void and may lead to immediate account deactivation. Abuse may result in the disabling of product features or even suspension of the account.

How to participate? Yupp participation process is as follows:

Sign up with your Google account (5,000 points for registration, and the official said that you can get an additional 2,500 points by using the yupp-launch code before June 20);

Ask questions to the AI model and choose the better answer between the two AI models;

Select the Feedback tab or provide text feedback.

Use your mouse to scratch the points scratch card to claim points.

You can decide whether to withdraw money by yourself (currently the official withdrawal has been temporarily closed).

ملخص As a16z crypto founder and managing partner Chris Dixon said, “Yupp’s design transforms human judgment into a sustainable economic resource. Data ‘expires’ as new interactions replace old data, creating a natural virtuous cycle: more usage leads to newer evaluations; newer evaluations lead to better models; better models attract more usage. All participants — from users to AI model builders — can participate and see that the same transparent rules apply to everyone, ensuring a credibly neutral market. No one can hide the leaderboard, and no one can manipulate rewards or outcomes.”

Yupps slogan is Every AI for everyone, which I think is more appropriate to translate as inclusive AI. Yupp attempts to build an evaluation infrastructure in the AI era through blockchain technology and crowdsourcing models: allowing users to get incentives through feedback, allowing developers to obtain real data, and ultimately promoting the evolution of AI technology in a more inclusive and trustworthy direction.

refer to:

https://www.wired.com/story/yupp-chatbot-pays-users-ai-model-feedback/

https://yupp.ai/help

https://blog.yupp.ai/launch

https://blog.yupp.ai/leaderboard

This article is sourced from the internet: Yupp raises $33 million in seed round, former Twitter technical backbone reconstructs AI model evaluation system

Related: KOL: How I made $100,000 by predicting market arbitrage Original author: Pix Original translation: Luffy, Foresight News Most people gamble on prediction markets, I arbitrage on prediction markets. Here is my specific strategy for making $100,000 from decentralized, inefficient prediction markets. Step 1: Understand the rules of the game Prediction markets allow you to bet on the outcomes of real-world events, such as: “Will Ethereum reach $5,000 by the end of the year?” “Will MrBeast run for president?” “Will Kanye West issue a token?” Each market has different user groups, each with their own biases. This means that the same event will be priced differently on different platforms, and this is where the opportunity lies. Example: If platform A’s “yes” quote is $0.4 and platform B’s “no” quote is $0.55, then no matter what the result is, you can…

تحليل #عملة # المشفرة# ديفيتبادل ## السوق© 版权声明المصفوفة 上一篇 BlockDAG Core dApp Guide: From BDAG Mobile Mining to Fan Tokens 下一篇 Trapped in Price Beliefs: The Psychological Black Hole of 90% of Crypto Investors 相关文章 CFTC’s new head takes the lead: How will Michael Selig Sworn balance crypto innovation and regulation? 6086cf14eb90bc67ca4fc62b 19٬002 3 توقعات البيتكوين لعام 2026: المخاطر والتنبؤات الكلية، والسوق يرسل إشارات حذرة 6086cf14eb90bc67ca4fc62b 31٬311 مراجعة إجراءات إنفاذ العملات المشفرة في الولايات المتحدة في السنوات الخمس الماضية: بلغ إجمالي أموال التسوية 1.10 تريليون و32 مليار دولار، مع FTX وAlame 6086cf14eb90bc67ca4fc62b 40٬167 2 The Truth Behind Bitcoin and Ethereum’s “Stagnation”: An Endogenous Deleveraging Dilemma 6086cf14eb90bc67ca4fc62b 7٬471 Hyperliquid’s market-making bot generates 220x profit in two weeks 6086cf14eb90bc67ca4fc62b 19٬654 2 Arthur Hayes: His heavy bet on ZEC was prompted by a wise person; the best investors have to think outside the box. 6086cf14eb90bc67ca4fc62b 17٬101 1 بدون تعليقات يجب عليك تسجيل الدخول لتترك تعليق! تسجيل الدخول على الفور بدون تعليقات... أحدث المقالات Did Jane Street “Manipulate” BTC? Decoding the AP System, Understanding the Power Struggle Behind ETF Creation and Redemption Pricing منذ 17 ساعة 542 Stop Comparing Bitcoin to Gold—It’s Now a High-Volatility Software Stock منذ 17 ساعة 659 Matrixport Research: $25 Billion Gamma Unwinding Imminent, Liquidity Yet to Return Behind the Rebound منذ 17 ساعة 608 ERC-5564: Ethereum’s Stealth Era Has Arrived, Receiving Addresses No Longer ‘Exposed’ منذ 17 ساعة 537 Hong Kong Regulatory Green Light: Asseto Enables DL Holdings to Achieve Compliance for Two RWA Business Implementations منذ 17 ساعة 598 المواقع الشعبيةTempoLighterGAIBطائرة شراعيةبلانكرايلزبوكر BCPokerفوي Bee.com أكبر بوابة Web3 في العالم الشركاء كوين كارب بينانس CoinMarketCap كوين جيكو كوين لايف الدروع قم بتنزيل تطبيق Bee Network وابدأ رحلة web3 ورق ابيض الأدوار التعليمات © 2021-2026. جميع الحقوق محفوظة. سياسة الخصوصية | شروط الخدمة تحميل تطبيق Bee Network وابدأ رحلة web3 أكبر بوابة Web3 في العالم الشركاء CoinCarp Binance CoinMarketCap CoinGecko Coinlive Armors ورق ابيض الأدوار التعليمات © 2021-2026. جميع الحقوق محفوظة. سياسة الخصوصية | شروط الخدمة يبحث يبحثفي الموقععلى تشيناجتماعيأخبار العنوان: صيادو الإنزال الجوي تحليل البيانات مشاهير التشفير كاشف الفخ العربية English 繁體中文 简体中文 日本語 Tiếng Việt 한국어 Bahasa Indonesia हिन्दी اردو Русский العربية

智能索引记录