Latest news with #PanguPro
Indian Express
08-07-2025
- Business
- Indian Express
Huawei's AI lab denies that one of its Pangu models copied Alibaba's Qwen
Huawei's artificial intelligence research division has rejected claims that a version of its Pangu Pro large language model has copied elements from an Alibaba model, saying that it was independently developed and trained. The division, called Noah Ark Lab, issued the statement on Saturday, a day after an entity called HonestAGI posted an English-language paper on code-sharing platform Github, saying Huawei's Pangu Pro Moe (Mixture of Experts) model showed 'extraordinary correlation' with Alibaba's Qwen 2.5 14B. This suggests that Huawei's model was derived through 'upcycling' and was not trained from scratch, the paper said, prompting widespread discussion in AI circles online and in Chinese tech-focused media. The paper added that its findings indicated potential copyright violation, the fabrication of information in technical reports and false claims about Huawei's investment in training the model. Noah Ark Lab said in its statement that the model was 'not based on incremental training of other manufacturers' models' and that it had 'made key innovations in architecture design and technical features.' It is the first large-scale model built entirely on Huawei's Ascend chips, it added. It also said that its development team had strictly adhered to open-source license requirements for any third-party code used, without elaborating which open-source models it took reference from. Alibaba did not immediately respond to a Reuters request for comment. Reuters was unable to contact HonestAGI or learn who is behind the entity. The release of Chinese startup DeepSeek's open-source model R1 in January this year shocked Silicon Valley with its low cost and sparked intense competition between China's tech giants to offer competitive products. Qwen 2.5-14B was released in May 2024 and is one of Alibaba's small-sized Qwen 2.5 model family which can be deployed on PC and smartphones. While Huawei entered the large language model arena early with its original Pangu release in 2021, it has since been perceived as lagging behind rivals. It open-sourced its Pangu Pro Moe models on Chinese developer platform GitCode in late June, seeking to boost the adoption of its AI tech by providing free access to developers. While Qwen is more consumer-facing and has chatbot services like ChatGPT, Huawei's Pangu models tend to be more used in government as well as the finance and manufacturing sectors.
United News of India
07-07-2025
- Business
- United News of India
Huawei denies claims that its Pangu model copied Alibaba's Qwen
New Delhi, July 7 (UNI) Huawei's AI research division on Monday outrightly denied the claims that the version of Pangu Pro large language models has copied some elements of Alibaba's Qwen. The division named 'Noah Ark Lab' said that Pangu is an independently trained model. Noah Ark Lab division has issued this statement just after the day HonestAGI posted a research paper on GitHub stating that Pangu Pro MoE (Mixture of Experts) has a deep correlation with Alibaba's Qwen 2.5-14B. The motive of this research paper is to show that the Pangu model is not trained from scratch and developed with the help of 'upcycling.' This revelation started a series of discussions in AI galore and Chinese tech-media. After the sensation spread, Noah Ark Lab said that the Pangu Model is not based on the incremental training of other manufacturers or models; it has made key innovations in architecture, design, and technical features. The statement also highlighted that the development team strictly followed open-source license requirements for any third-party code used. UNI SAS BM More News Siemens Energy India Ltd Posts Q2 FY26 Net Profit Of Rs 246 Crore 07 Jul 2025 | 6:33 PM Mumbai, July 7 (UNI) Siemens Energy India Ltd (SEIL) officially announced here on Monday that it has posted a net profit of Rs 246.1 crore for the January-March 2025 period (Q2 FY26), which is a 36.34% increase as compared to the corresponding period during the year-ago period. see more.. Huawei denies claims that its Pangu model copied Alibaba's Qwen 07 Jul 2025 | 6:10 PM New Delhi, July 7 (UNI) Huawei's AI research division on Monday outrightly denied the claims that the version of Pangu Pro large language models has copied some elements of Alibaba's Qwen. see more.. Nigeria Data Protection Commission imposes a fine on MultiChoice Nigeria 07 Jul 2025 | 5:29 PM New Delhi, July 7 (UNI) In the latest tech developments, NDPC (Nigeria Data Protection Commission) has imposed a fine on MultiChoice Nigeria, which is a joint venture. see more.. TVS launches Jupiter 110 scooter in Nepal market 07 Jul 2025 | 5:22 PM Chennai, July 7 (UNI) Two and three wheeler makers TVS Motor Company today launched the all new TVS Jupiter 110 in the Nepal market. see more.. Kottayam-Kochi Rubber Market Rates 07 Jul 2025 | 5:17 PM Kottayam, July 7 (UNI) Following were the Rubber Market rates announced by the Rubber Board here today per quintal KOTTAYAM RSS FOUR: 20300 RSS FIVE : 20000 ISNR TWENTY: 18200 SIXTY PERCENT LATEX: 14160 KOCHI RSS FOUR : 20300 RSS FIVE : 20000 INTERNATIONAL RATES (BANGKOK) RSS One : 19595 RSS Two : 19437 RSS Three: 19292 RSS Four : 19213 RSS Five : 19094 KUALALUMPUR SMR TWENTY : 14392 ($168.55) SIXTY PERCENT LATEX : 11627 ($136.15) UNI PA BM. see more..
Yahoo
07-07-2025
- Business
- Yahoo
Huawei's AI Scandal Just Exploded--And Investors Should Be Paying Attention
Huawei is pushing backhard. Over the weekend, its secretive Noah's Ark Lab broke from its usual silence to address accusations that its new AI model, Pangu Pro MoE, borrowed code without proper credit. The model, which runs on Huawei's own Ascend chips (their homegrown answer to Nvidia's GPUs), had its source code picked apart on GitHub, where a group dubbed HonestAGI claimed it spotted unacknowledged code fragments. That post vanished. But another one, titled Pangu's Sorrow, quickly followedalleging that Huawei's team had been under intense pressure to deliver and fell behind domestic rivals in the race. In a rare rebuttal, Huawei said it fully complied with open-source licenses and welcomed technical discussion, not speculation. This isn't just a code reviewit's a window into the internal pressure mounting inside China's AI champions. With Alibaba and DeepSeek making waves and catching investor attention, Huawei's rare public statement signals how high the stakes are becoming. The companylong a symbol of China's tech self-sufficiencynow finds itself on the defensive in one of the hottest battlegrounds: sovereign AI. The fact that Huawei had to respond at all speaks volumes. IP compliance, innovation speed, and trust are no longer soft issuesthey're table stakes in an environment where reputations are earned (or lost) in public. For global investors watching the AI value chainfrom chipmakers like Nvidia (NASDAQ:NVDA) to downstream platforms like Tesla (NASDAQ:TSLA)this is another flashing signal. The game isn't just about who builds the fastest model. It's also about who's playing fair, who's shipping on time, and who's earning credibility in a world that increasingly demands transparency. As China's AI ecosystem matures, these reputational battles could become just as important as the hardware wars. This article first appeared on GuruFocus. Error while retrieving data Sign in to access your portfolio Error while retrieving data Error while retrieving data Error while retrieving data Error while retrieving data
Yahoo
07-07-2025
- Business
- Yahoo
Huawei's AI Scandal Just Exploded--And Investors Should Be Paying Attention
Huawei is pushing backhard. Over the weekend, its secretive Noah's Ark Lab broke from its usual silence to address accusations that its new AI model, Pangu Pro MoE, borrowed code without proper credit. The model, which runs on Huawei's own Ascend chips (their homegrown answer to Nvidia's GPUs), had its source code picked apart on GitHub, where a group dubbed HonestAGI claimed it spotted unacknowledged code fragments. That post vanished. But another one, titled Pangu's Sorrow, quickly followedalleging that Huawei's team had been under intense pressure to deliver and fell behind domestic rivals in the race. In a rare rebuttal, Huawei said it fully complied with open-source licenses and welcomed technical discussion, not speculation. This isn't just a code reviewit's a window into the internal pressure mounting inside China's AI champions. With Alibaba and DeepSeek making waves and catching investor attention, Huawei's rare public statement signals how high the stakes are becoming. The companylong a symbol of China's tech self-sufficiencynow finds itself on the defensive in one of the hottest battlegrounds: sovereign AI. The fact that Huawei had to respond at all speaks volumes. IP compliance, innovation speed, and trust are no longer soft issuesthey're table stakes in an environment where reputations are earned (or lost) in public. For global investors watching the AI value chainfrom chipmakers like Nvidia (NASDAQ:NVDA) to downstream platforms like Tesla (NASDAQ:TSLA)this is another flashing signal. The game isn't just about who builds the fastest model. It's also about who's playing fair, who's shipping on time, and who's earning credibility in a world that increasingly demands transparency. As China's AI ecosystem matures, these reputational battles could become just as important as the hardware wars. This article first appeared on GuruFocus.
Time of India
07-07-2025
- Business
- Time of India
Huawei's AI lab denies that one of its Pangu models copied Alibaba's Qwen
BEIJING/SHANGHAI: Huawei 's artificial intelligence research division has rejected claims that a version of its Pangu Pro large language model has copied elements from an Alibaba model, saying that it was independently developed and trained. The division, called Noah Ark Lab , issued the statement on Saturday, a day after an entity called HonestAGI posted an English-language paper on code-sharing platform Github, saying Huawei's Pangu Pro Moe (Mixture of Experts) model showed "extraordinary correlation" with Alibaba's Qwen 2.5 14B. This suggests that Huawei's model was derived through "upcycling" and was not trained from scratch, the paper said, prompting widespread discussion in AI circles online and in Chinese tech-focused media. The paper added that its findings indicated potential copyright violation, the fabrication of information in technical reports and false claims about Huawei's investment in training the model. Noah Ark Lab said in its statement that the model was "not based on incremental training of other manufacturers' models" and that it had "made key innovations in architecture design and technical features." It is the first large-scale model built entirely on Huawei's Ascend chips, it added. It also said that its development team had strictly adhered to open-source license requirements for any third-party code used, without elaborating which open-source models it took reference from. Alibaba did not immediately respond to a Reuters request for comment. Reuters was unable to contact HonestAGI or learn who is behind the entity. The release of Chinese startup DeepSeek's open-source model R1 in January this year shocked Silicon Valley with its low cost and sparked intense competition between China's tech giants to offer competitive products. Qwen 2.5-14B was released in May 2024 and is one of Alibaba's small-sized Qwen 2.5 model family which can be deployed on PC and smartphones. While Huawei entered the large language model arena early with its original Pangu release in 2021, it has since been perceived as lagging behind rivals. It open-sourced its Pangu Pro Moe models on Chinese developer platform GitCode in late June, seeking to boost the adoption of its AI tech by providing free access to developers. While Qwen is more consumer-facing and has chatbot services like ChatGPT, Huawei's Pangu models tend to be more used in government as well as the finance and manufacturing sectors.
