Aleph Alpha

𝗰𝗿𝗲𝗮𝗻𝗰𝗲.𝗮𝗶, a joint venture between PwC Deutschland and Aleph Alpha, yesterday kicked off a pilot project with Deutsche Bank, Boerse Stuttgart Group, and VÖB-Service GmbH at PwC's Frankfurt office! Together, we will evaluate the potential of #GenAI for AI-driven compliance solutions. A big thank you to our team and our partners!

𝗰𝗿𝗲𝗮𝗻𝗰𝗲.𝗮𝗶 launches pilot project with Deutsche Bank, Börse Stuttgart Group, and VÖB Service   We’re excited to share that #creanceai, a joint venture between PwC Deutschland and Aleph Alpha, has initiated a pilot project with the leading financial institutions Deutsche Bank, Boerse Stuttgart Group and VÖB-Service GmbH 𝗰𝗿𝗲𝗮𝗻𝗰𝗲.𝗮𝗶 combines Aleph Alpha’s first-class #AI technology with PwC’s in-depth, global expertise. Deutsche Bank, Gruppe Börse Stuttgart, and VÖB-Service will work with 𝗰𝗿𝗲𝗮𝗻𝗰𝗲.𝗮𝗶 to evaluate how generative AI empowers experts to master complex #compliance challenges. Discover how 𝗰𝗿𝗲𝗮𝗻𝗰𝗲.𝗮𝗶 is shaping the future of compliance with #GenAI technology for human-machine compliance verifications, workflows, and improvements! #writtenbyalephalpha

𝗔𝗹𝗲𝗽𝗵 𝗔𝗹𝗽𝗵𝗮 𝘀𝗶𝗴𝗻𝘀 𝘁𝗵𝗲 𝗔𝗜 𝗣𝗮𝗰𝘁 ✅ Only secure and trustworthy GenAI solutions create sustainable value for enterprises and public administrations. A unified European framework promotes, secures and creates an environment for responsible innovation. In addition to our scientific achievements and unique technical features in the areas of explainability and transparency, we have already created an unparalleled level of compliance and security for our customers with #PhariaAI. For us, this establishes the decisive foundation for sovereignty. #EuropeanAIPact 𝗔𝗹𝗲𝗽𝗵 𝗔𝗹𝗽𝗵𝗮 𝘂𝗻𝘁𝗲𝗿𝘇𝗲𝗶𝗰𝗵𝗻𝗲𝘁 𝗱𝗲𝗻 𝗔𝗜 𝗣𝗮𝗰𝘁 ✅ Nur sichere und vertrauenswürdige GenAI-Lösungen schaffen nachhaltige Werte für Unternehmen und öffentliche Verwaltungen. Ein einheitlicher europäischer Rahmen fördert, sichert und etabliert ein Umfeld für Innovation mit Verantwortung. Neben unseren akademischen Erfolgen und technischen Alleinstellungsmerkmalen im Bereich Erklärbarkeit und Transparenz, haben wir mit #PhariaAI für unsere Kunden bereits heute ein einzigartiges Niveau an Compliance and Sicherheit geschaffen. Damit entsteht für uns das entscheidende Fundament für Souveränität.    #EuropeanAIPact

Gemeinsam mit PwC Deutschland freuen wir uns auf die Creance.ai Veranstaltung am kommenden Montag in Frankfurt. Mehr dazu hier: https://lnkd.in/eUyRUT_H

Am 30. September 2024 ist es soweit: Im PwC Experience Center in Frankfurt dreht sich alles um die Zukunft von GenAI und Compliance. Wir freuen uns, wenn Sie dabei sind.

Introducing T-Free, a new paradigm for language processing Aleph Alpha Research is proud to introduce T-Free, a new paradigm in language processing. Tokenization is one of the core building blocks of large language models (LLMs), transforming natural language into numeric representations for further processing. Our innovation, T-Free, offers a novel approach to tokenization, boosting tokenizer fertility across various languages, and reducing the size of the embedding layer by up to 75% compared to traditional tokenizers.  Early experiments with T-Free show promising results and could unlock new possibilities in LLMs, including: - Up to 50% reduction in training and inference costs  - Improved semantic encoding of language  - Enhanced performance in multilingual models 🔗 To learn more about T-Free, check out our technical blog post: https://lnkd.in/eTi7kjuc #writtenbyalephalpha #Lab1141

Today we launched 𝗣𝗵𝗮𝗿𝗶𝗮𝗔𝗜 - our new 𝗲𝗻𝗱-𝘁𝗼-𝗲𝗻𝗱 𝗔𝗜 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻 - as part of our strategy update together with our partners and customers.  𝗣𝗵𝗮𝗿𝗶𝗮𝗔𝗜 is a full stack offer for enterprises and governments, built to develop, deploy and distribute sovereign generative AI technology. 𝗣𝗵𝗮𝗿𝗶𝗮𝗔𝗜 combines proprietary innovation with our new state-of-the-art open-source models and components and is meeting the requirements of the most complex and critical information-based value chains even in highly regulated environments. 𝗣𝗵𝗮𝗿𝗶𝗮𝗔𝗜 features our new model Pharia-1-LLM which we just launched this morning. Learn more about 𝗣𝗵𝗮𝗿𝗶𝗮𝗔𝗜! https://lnkd.in/e739KRHp #writtenbyalephalpha

𝗣𝗵𝗮𝗿𝗶𝗮-𝟭-𝗟𝗟𝗠 𝗶𝘀 𝗵𝗲𝗿𝗲!   We are excited to launch 𝗣𝗵𝗮𝗿𝗶𝗮-𝟭-𝗟𝗟𝗠 with two 7B foundation models: 𝗣𝗵𝗮𝗿𝗶𝗮-𝟭-𝗟𝗟𝗠-𝟳𝗕-𝗰𝗼𝗻𝘁𝗿𝗼𝗹 and 𝗣𝗵𝗮𝗿𝗶𝗮-𝟭-𝗟𝗟𝗠-𝟳𝗕-𝗰𝗼𝗻𝘁𝗿𝗼𝗹-𝗮𝗹𝗶𝗴𝗻𝗲𝗱. Both models and the code used to train them are now publicly available and open-sourced for non-commercial and educational use. 𝗣𝗵𝗮𝗿𝗶𝗮-𝟭-𝗟𝗟𝗠 models are trained on 7 languages and culturally and linguistically optimized for English, German, French and Spanish, tailored towards concise responses, and stand on par with the latest open-source models in the 7B to 8B parameter range. They have been trained in full compliance with applicable EU and national regulations, including copyright and data privacy laws. Following our commitment to transparency we have decided to go one step further than most industrial AI R&D labs by making not only our models but also the code base used for training the models openly available for non-commercial use. Find more info on our models & training code base in our model card or dive deep into the technical details in the two accompanying blog posts.    https://lnkd.in/ejEeyVeZ #writtenbyalephalpha

Aleph Alpha x TU Darmstadt    We're thrilled to announce Lab 1141, a groundbreaking collaboration between Aleph Alpha Research and Technische Universität Darmstadt, with the goal of pushing the boundaries of explainable, safe, and transparent GenAI.     Lab 1141 will close the gap between academic research and industrial innovation, advancing AI transparency and the AI explainability research by contributing to the open research community while driving product innovation.    Lab 1141 provides a collaborative environment in which PhD students and Post-Docs at TU Darmstadt and Aleph Alpha Research are funded through Aleph Alpha Research Fellowships. The setup fosters mechanisms of knowledge exchange and collaborative research between academia and industry to translate scientific excellence into innovative, cutting-edge real-world applications.     We invite brilliant PhD candidates to apply to the upcoming positions linked in our profile. 🔗 https://lnkd.in/eWyGU-sE   #writtenbyalephalpha #lab1141

Unser CEO Jonas Andrulis im Gespräch mit unserem Partner Materna Information & Communications SE CEO Michael Hagedorn darüber, ➡ an welchen Stellen und wie souveräne KI 𝗠𝗲𝗵𝗿𝘄𝗲𝗿𝘁 🎯 für Unternehmen schafft und ➡ warum die 𝗿𝗮𝗽𝗶𝗱𝗲𝗻 𝗙𝗼𝗿𝘁𝘀𝗰𝗵𝗿𝗶𝘁𝘁𝗲 in KI 𝘀𝘂𝗯𝘀𝘁𝗮𝗻𝘁𝗶𝗲𝗹𝗹𝗲 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻❗der Wissensarbeit in Unternehmen und Europa bedeuten. Danke für die starke, zukunftsweisende Partnerschaft Materna Information & Communications SE #GenAI #SovereignAI #AItransformation #AIpartnership