current location：Home > Technical Articles > Technology peripherals > AI
- Direction：
- All web3.0 Backend Development Web Front-end Database Operation and Maintenance Development Tools PHP Framework Daily Programming WeChat Applet Common Problem Other Tech CMS Tutorial Java System Tutorial Computer Tutorials Hardware Tutorial Mobile Tutorial Software Tutorial Mobile Game Tutorial
- Classify：
-
- Add fast and slow eyes to the video model, Apple's new training-free method surpasses everything SOTA in seconds
- Since the release of Sora, the field of AI video generation has become more "busy". In the past few months, we have witnessed Jimeng, RunwayGen-3, LumaAI, and Kuaishou Keling taking turns to explode. Unlike in the past, where one could tell at a glance that the models were generated by AI, this batch of large video models may be the “best” we have ever seen. However, the amazing performance of video large language models (LLM) is inseparable from a large and finely annotated video data set, which requires a very high cost. Recently, a number of innovative methods have emerged in the research field that do not require additional training: using trained image large language models to directly process video tasks, thus bypassing the "expensive" training process. In addition, most existing video LLMs
- AI 601 2024-08-11 16:02:31
-
- Hybrid experts are more assertive and can perceive multiple modalities and act according to the situation. Meta proposes modality-aware expert hybrids
- Mixed experts also have specializations in their profession. For current mixed-modality basic models, a common architectural design is to fuse encoders or decoders of specific modalities. However, this method has limitations: it cannot integrate information from different modalities, and it is difficult to output content containing multiple modalities. . In order to overcome this limitation, MetaFAIR's Chameleon team proposed a new single Transformer architecture in the recent paper "Chameleon: Mixed-modalearly-fusionfoundationmodels", which can transform discrete images and text based on the prediction target of the next token. The mixed modal sequence composed of tokens is modeled, so that in different modes
- AI 488 2024-08-11 13:02:22
-
- Open and closed source model 'Chaos': Let's see which agent can best glimpse human beings' true intentions
- The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com The first authors of this article are Qian Cheng and He Bingxiang, undergraduate students in the Department of Computer Science at Tsinghua University. Both are members of THUNLP. Qian Cheng's main research interests are tool learning and large model-driven agents. He is about to study for a PhD at UIUC. He Bingxiang’s main research interests are large model alignment and
- AI 1300 2024-08-11 13:02:02
-
- The first artificial intelligence party was successfully held
- In 1905, the first film "Dingjun Mountain" shot by the Chinese themselves was born in Beijing Fengtai Photo Studio, ending the history of no domestic films in China. More than a hundred years later, on August 28, 2023, in the "Original Heart" Hall of the People's Daily New Media Building, a song "New Dingjun Mountain" was performed in "cross-dimensional cooperation" by college students from the Chinese Academy of Opera and digital person Mu Lan , triggering bursts of applause from the audience present. The ancient Peking Opera art, iterative light and shadow technology and new digital technology pay tribute to the classics while breaking through the times and showing a new charm. This is one of the major programs launched by the first Artificial Intelligence (AI) Gala hosted by People's Daily Online. Through a series of cultural festivals full of technology, intelligence and future, the party
- AI 594 2024-08-10 22:23:02
-
- 2024 World Artificial Intelligence Conference TOP20 Young Excellent Papers Released
- In March 2024, the "Notice on Recommending Participating Papers for the "2024 World Artificial Intelligence Conference Youth Outstanding Paper Award"" was released, and a collection activity for young outstanding papers in the field of artificial intelligence was launched for universities, scientific research institutes, and enterprises around the world. As of the deadline for submissions, a total of 159 review papers have been received from domestic and foreign countries, including internationally renowned universities, scientific research institutions, and enterprises. After preliminary evaluation and re-evaluation, the TOP20 outstanding youth papers in 2024 are now released. (In alphabetical order by paper field and name) The 2024 World Artificial Intelligence Conference Youth Outstanding Paper Award will be announced at the World Artificial Intelligence Conference in July, so stay tuned!
- AI 1042 2024-08-10 16:06:11
-
- Forum Preview | 'Knowledge Changes the World - Vertical Large Model Redefines Knowledge Management' forum is about to open
- Introduction: Large model technology has great potential in accelerating knowledge acquisition, integration, application and innovation, and can bring transformative impacts to many industries such as education, finance, and government affairs. At present, the application of vertical large models in enterprise scenarios in my country is still in the exploratory stage. Many enterprises are still seeking the best implementation plan for large models. Among them, deep integration with the enterprise's internal knowledge base is regarded as the enterprise-level scenario for the implementation of vertical large models. important direction of application. To this end, during the 2024 World Artificial Intelligence Conference and the High-level Conference on Artificial Intelligence Global Governance, Daguan Data will join hands with Fudan-Daguan Financial Vertical Application Large Model School-Enterprise Joint Research Center and Pudong Youth Federation to jointly host "Knowledge Changes the World-Vertical Big Forum on the topic "Model Redefines Knowledge Management". The forum will be held in July 2024
- AI 1018 2024-08-10 11:20:06
-
- Selected into ICDE, the top international database conference, Tencent Cloud database technology innovation has been recognized by the authority
- I learned today that two papers from Tencent Cloud Database were selected into ICDE, the top international database conference. Its distributed transaction protocol - Lion and memory-disk spanning index design framework have solved common problems in the database field, and its technological innovation has been recognized by international authorities. Tencent Cloud Database Team shared technology at the ICDE venue. ICDE (IEEE International Conference on Data Engineering) is an international conference in the field of database research. Together with SIGMOD and VLDB, it is also known as the three top database conferences. Distributed database transaction processing Distributed transaction processing involves multiple rounds of cross-node communication and is slow. past migration
- AI 731 2024-08-10 07:10:06
-
- AI Helps Me With My Schoolwork, But Doesn\'t Do It for Me (ChatGPT, NotebookLM)
- I am a straight-A graduate student with mixed feelings on AI tools. Many people use AI to replace their own efforts in an academic setting. But you can use AI to assist your learning, add efficiency to research, and improve your studies. Start In-De
- AI 838 2024-08-10 06:43:32
-
- Tongyi Qianwen open sourced Qwen2-Math, becoming the most advanced mathematics-specific model
- According to news on August 9, Alibaba Tongyi team open sourced a new generation mathematical model Qwen2-Math, which includes a basic model and an instruction fine-tuning model with three parameters of 1.5B, 7B, and 72B. Qwen2-Math is developed based on the Tongyi Qianwen open source large language model Qwen2. The flagship model Qwen2-Math-72B-Instruct scores higher than GPT-4o, Claude-3.5-Sonnet, Gemini-1.5-Pro, and Llama on the authoritative evaluation set MATH. -3.1-405B, etc., handle a variety of mathematical problems such as algebra, geometry, counting and probability, number theory, etc. with an accuracy of 84%, becoming the most advanced mathematics-specific model. NOTE: In MATH benchmark evaluation
- AI 938 2024-08-09 19:44:14
-
- Google I/O Developer Conference China Station: Large model technology enters the entire ecosystem
- On Wednesday, Google held its China Developers Conference in Beijing. In addition to the I/O Conference in Mountain View, California in May, this year's I/OConnect global tour started from Berlin, Germany, to Bangalore, India, and has also arrived in Beijing, China. At this event, experts from many different fields shared in-depth Google's latest technological progress, development tools, and global service platforms in AI, Web, mobile, cloud technology and other fields for domestic developers. "China is one of the largest developer markets in the world, and our developers have always been an indispensable pioneering force on the global stage," Chen Junting, president of Google Greater China, said in the opening speech. "In the past year, 25 development teams from China have produced a total of 31 games and applications in different regions around the world.
- AI 673 2024-08-09 19:42:02
-
- Complete 1 year MD calculation in 2.5 days? DeepMind team's new calculation method based on Euclidean Transformer
- Editor | Radish Skin In recent years, great progress has been made in the development of machine learning force fields (MLFF) based on ab initio reference calculations. Although low test errors are achieved, the reliability of MLFF in molecular dynamics (MD) simulations is facing increasing scrutiny due to concerns about instability over longer simulation time scales. Research has shown a potential link between robustness to cumulative inaccuracies and the use of equivariant representations in MLFF, but the computational costs associated with these representations may limit this advantage in practice. To solve this problem, researchers from Google DeepMind and Technical University of Berlin (TUBerlin) proposed a transformer framework called SO3krates.
- AI 773 2024-08-09 19:40:30
-
- Karpathy's views are controversial: RLHF is not real reinforcement learning, and Google and Meta are opposed to it
- It seems that everyone still has different opinions on whether RLHF and RL can be classified into the same category. AI guru Karpathy is here again to popularize the concept of artificial intelligence. Yesterday, he tweeted, "Reinforcement learning based on human feedback (RLHF) is just reinforcement learning (RL)." Karpathy's full text is explained as follows: RLHF is the third (and also the third) method for training large language models (LLM). The last) main stage, the first two stages are pre-training and supervised fine-tuning (SFT). I think RLHF is just barely considered RL and it is not widely recognized. RL is powerful, but RLHF is not. Let's look at the example of AlphaGo, which was trained using real RL. computer games
- AI 1080 2024-08-09 16:40:09
-
- Two papers were nominated for the Best Paper Honorable Mention at the same time. The first Real-Time Live Chinese team at SIGGRAPH uses generative AI to create a 3D world.
- SIGGRAPH, the top global academic conference focusing on computer graphics, is emerging with a new trend. At the SIGGRAPH2024 conference held last week, among the best paper and other awards, the team from the MARS Laboratory of Shanghai University of Science and Technology received two honorary nominations for the best paper at the same time, and its research results are also rapidly moving toward industrialization. The author uses the method of generative models to open up a new way to directly transform imagination into complex 3D models. The two papers nominated for the best paper - CLAY and DressCode, are about 3D generation and 3D clothing generation respectively. In the Real-TimeLive session of SIGGARPH, the team from Shanghai University of Science and Technology demonstrated real-time results based on these two works.
- AI 1080 2024-08-09 16:03:49
-
- DeepMind robot plays table tennis, and its forehand and backhand slip into the air, completely defeating human beginners
- But maybe he can’t defeat the old man in the park? The Paris Olympic Games are in full swing, and table tennis has attracted much attention. At the same time, robots have also made new breakthroughs in playing table tennis. Just now, DeepMind proposed the first learning robot agent that can reach the level of human amateur players in competitive table tennis. Paper address: https://arxiv.org/pdf/2408.03906 How good is the DeepMind robot at playing table tennis? Probably on par with human amateur players: both forehand and backhand: the opponent uses a variety of playing styles, and the robot can also withstand: receiving serves with different spins: However, the intensity of the game does not seem to be as intense as the old man in the park. For robots, table tennis
- AI 1450 2024-08-09 16:01:32
-
- Will speculative sampling lose the inference accuracy of large language models?
- The prototype concept of speculative sampling was proposed by MitchellStern et al. in 2018. This approach has since been further developed and refined by various works, including LookaheadDecoding, REST, Medusa, and EAGLE, where speculative sampling significantly speeds up the inference process of large language models (LLMs). An important question is: does speculative sampling in LLM harm the accuracy of the original model? Let me start with the answer: no. The standard speculative sampling algorithm is lossless, and this article will prove this through mathematical analysis and experiments. Mathematically, the speculative sampling formula can be defined as follows: where: ? is a real number sampled from a uniform distribution. is the next token to be predicted. ?(?) is given by the draft model
- AI 1273 2024-08-09 13:09:05
Tool Recommendations
jQuery enterprise message form contact code
jQuery enterprise message form contact code is a simple and practical enterprise message form and contact us introduction page code.
form button
2024-02-29
HTML5 MP3 music box playback effects
HTML5 MP3 music box playback special effect is an mp3 music player based on HTML5 css3 to create cute music box emoticons and click the switch button.
Player special effects
2024-02-29
HTML5 cool particle animation navigation menu special effects
HTML5 cool particle animation navigation menu special effect is a special effect that changes color when the navigation menu is hovered by the mouse.
Menu navigation
2024-02-29
jQuery visual form drag and drop editing code
jQuery visual form drag and drop editing code is a visual form based on jQuery and bootstrap framework.
form button
2024-02-29
Organic fruit and vegetable supplier web template Bootstrap5
An organic fruit and vegetable supplier web template-Bootstrap5
Bootstrap template
2023-02-03
Bootstrap3 multifunctional data information background management responsive web page template-Novus
Bootstrap3 multifunctional data information background management responsive web page template-Novus
backend template
2023-02-02
Real estate resource service platform web page template Bootstrap5
Real estate resource service platform web page template Bootstrap5
Bootstrap template
2023-02-02
Simple resume information web template Bootstrap4
Simple resume information web template Bootstrap4
Bootstrap template
2023-02-02
Cute summer elements vector material (EPS PNG)
This is a cute summer element vector material, including the sun, sun hat, coconut tree, bikini, airplane, watermelon, ice cream, ice cream, cold drink, swimming ring, flip-flops, pineapple, conch, shell, starfish, crab, Lemons, sunscreen, sunglasses, etc., the materials are provided in EPS and PNG formats, including JPG previews.
PNG material
2024-05-09
Four red 2023 graduation badges vector material (AI EPS PNG)
This is a red 2023 graduation badge vector material, four in total, available in AI, EPS and PNG formats, including JPG preview.
PNG material
2024-02-29
Singing bird and cart filled with flowers design spring banner vector material (AI EPS)
This is a spring banner vector material designed with singing birds and a cart full of flowers. It is available in AI and EPS formats, including JPG preview.
banner picture
2024-02-29
Golden graduation cap vector material (EPS PNG)
This is a golden graduation cap vector material, available in EPS and PNG formats, including JPG preview.
PNG material
2024-02-27
Home Decor Cleaning and Repair Service Company Website Template
Home Decoration Cleaning and Maintenance Service Company Website Template is a website template download suitable for promotional websites that provide home decoration, cleaning, maintenance and other service organizations. Tip: This template calls the Google font library, and the page may open slowly.
Front-end template
2024-05-09
Fresh color personal resume guide page template
Fresh color matching personal job application resume guide page template is a personal job search resume work display guide page web template download suitable for fresh color matching style. Tip: This template calls the Google font library, and the page may open slowly.
Front-end template
2024-02-29
Designer Creative Job Resume Web Template
Designer Creative Job Resume Web Template is a downloadable web template for personal job resume display suitable for various designer positions. Tip: This template calls the Google font library, and the page may open slowly.
Front-end template
2024-02-28
Modern engineering construction company website template
The modern engineering and construction company website template is a downloadable website template suitable for promotion of the engineering and construction service industry. Tip: This template calls the Google font library, and the page may open slowly.
Front-end template
2024-02-28
