Rain Neuromorphics is involved in developing neuromorphic hardware, which is a cutting-edge technology in the field of artificial intelligence. Their Memristive Nanowire Neural Network (MN3) is described as a state-of-the-art model that uses unique materials and techniques to create a highly efficient and advanced AI processor. This aligns with the criteria of developing frontier models that are considered cutting-edge or state-of-the-art in their field.
Layer 6 AI is involved in developing advanced machine learning and deep learning systems, which are considered cutting-edge and state-of-the-art in their field. Their work spans multiple domains, including healthcare, finance, and recommendation systems, and they have been recognized for their technical capabilities through international machine learning competition wins. This aligns with the criteria of developing frontier models.
Geminus.ai specializes in physics-informed AI and combines physics-based simulations with deep learning for real-time operational decisions. Their products include autonomous systems with self-optimizing and self-healing capabilities, and they offer model predictive control with evergreen, purpose-built AI models. These features suggest that they are developing cutting-edge, state-of-the-art models in their field, which aligns with the criteria of building frontier models.
Cerebras Systems is involved in developing frontier models as evidenced by their AI supercomputers, AI model services, and the CS-3 system, which is described as the world’s fastest and most scalable AI accelerator. Their technology is considered cutting-edge and state-of-the-art, particularly in the field of AI and high-performance computing.
Hume AI is involved in developing frontier models focused on emotional intelligence in AI, which are considered cutting-edge in their field. The company conducts extensive research, has published numerous papers, and has developed advanced APIs for interpreting vocal and facial expressions, aligning with the criteria of building state-of-the-art frontier models.
Deepgram is involved in developing advanced speech AI transcription and understanding models, which can be considered cutting-edge in the field of Natural Language Understanding (NLU). The company offers features like summarization, sentiment analysis, and topic detection, and allows for the training of custom models for unique use cases. Additionally, Deepgram's CTO and VP of Research have strong backgrounds in machine learning and deep learning, further supporting the company's capability in building state-of-the-art models.
AssemblyAI is involved in developing frontier models, specifically in the field of Speech AI. They have developed state-of-the-art models like Conformer-2, which is trained on 1.1M hours of voice data and makes up to 43% fewer errors on noisy data compared to other models. They are also working on a next-gen Universal model trained on over 10M hours of voice data using Google's TPU chips. These models are considered cutting-edge in their field.
Normal Computing is involved in developing advanced AI models, including generative AI for critical enterprise applications and advanced reasoning and planning algorithms with LLMs. The company also focuses on cutting-edge research such as thermodynamic matrix inversion and eliminating hallucinations in large language models, indicating that their work is state-of-the-art in the field.
Atomic AI is involved in developing frontier models as it utilizes deep learning foundation models and has created the first large language model using chemical mapping data for RNA therapeutic development. Their technology is considered cutting-edge, as evidenced by being featured on the cover of Science and their focus on previously untreatable diseases.
Cancel
