Latest news with #Sapient
Associated Press
21-07-2025
- Business
- Associated Press
Sapient Intelligence Open-Sources Hierarchical Reasoning Model, a Brain-Inspired Architecture That Solves Complex Reasoning Tasks With 27 Million Parameters
A 27 M-parameter, brain-inspired architecture cracks ARC-AGI, Sudoku-Extreme, and Maze-Hard with just 1000 training examples and without pre-training Singapore - 21 July, 2025 - AGI Research Company Sapient Intelligence today announced the open-source release of its Hierarchical Reasoning Model (HRM), a brain-inspired architecture that leverages hierarchical structure and multi-timescale processing to achieve substantial computational depth without sacrificing training stability or efficiency. Trained on just 1000 examples without pre-training, with only 27 million parameters, HRM successfully tackles reasoning challenges that continue to frustrate today's large language models (LLMs). Beyond LLMs' Reasoning Limits Current LLMs depend heavily on Chain-of-Thought prompting, an approach that often suffers from brittle task decomposition, immense training data demands and high latency. Inspired by the hierarchical and multi-timescale processing in the human brain, HRM overcomes these constraints by embracing three fundamental principles observed in cortical computation: hierarchical processing, temporal separation, and recurrent connectivity. Composed of a high-level module performing slow, abstract planning and a low-level module executing rapid, detailed computations, HRM is capable of alternating dynamically between automatic thinking ('System 1') and deliberate reasoning ('System 2') in a single forward pass. 'AGI is really about giving machines human-level, and eventually beyond-human, intelligence. CoT lets the models imitate human reasoning by playing the odds, and it's only a workaround. At Sapient, we're starting from scratch with a brain-inspired architecture, because nature has already spent billions of years perfecting it. Our model actually thinks and reasons like a person, not just crunches probabilities to ace benchmarks. We believe it will reach, then surpass, human intelligence, and that's when the AGI conversation gets real,' said Guan Wang, founder and CEO of Sapient Intelligence. Inspired by the brain, HRM has two recurrent networks operating at different timescales to collaboratively solve tasks Benchmark Breakthroughs Despite its compact scale of 27 million parameters and using only 1000 input-output examples, all without any pre-training or Chain-of-Thought supervision, HRM learns to solve problems that even the most advanced LLMs struggle with. In the Abstraction and Reasoning Corpus (ARC) AGI Challenge, a widely accepted benchmark of inductive reasoning, HRM archives a performance of 5% on ARC-AGI-2, significantly outperforming OpenAI o3-mini-high, DeepSeek R1, and Claude 3.7 8K, all of which rely on far larger sizes and context lengths. In complex Sudoku puzzles and optimal pathfinding in 30x30 mazes, where state-of-the-art CoT methods completely fail, HRM delivers near-perfect accuracy. With only about 1000 training examples, the HRM (~27M parameters) surpasses state-of-the-art CoT models on ARC-AGI, Sudoku-Extreme, and Maze-Hard* The Sapient Intelligence team is already running new experiments and expects to publish even stronger ARC-AGI scores soon. Real-World Impact HRM's data efficiency and reasoning accuracy open new opportunities in fields where large datasets are scarce yet accuracy is critical. In healthcare, Sapient is partnering with leading medical research institutions to deploy HRM to support complex diagnostics, particularly rare-disease cases where data signals are sparse, subtle, and demand deep reasoning. In climate forecasting, HRM raises subseasonal-to-seasonal (S2S) forecasting accuracy to 97 %, a leap that translates directly into social and economic value. In robotics, HRM's low-latency, lightweight architecture serves as an on-device 'decision brain,' enabling next-generation robots to perceive and act in real time within dynamic environments. Path Forward Sapient Intelligence believes that HRM presents a viable alternative to the currently dominant CoT reasoning models. It offers a practical path toward universally capable reasoning systems that rely on architecture, not scale, to push the frontier of AI and, ultimately, close the gap between today's models and true artificial general intelligence. Availability The source code is available on GitHub at About Sapient Intelligence Sapient Intelligence is a global AGI research company headquartered in Singapore, with research centers in San Francisco and Beijing, building the next-generation AI model for complex reasoning. Our mission is to reach artificial general intelligence by developing a radically new architecture that integrates reinforcement learning, evolutionary algorithms, and neuroscience research to push beyond the limits of today's LLMs. In July 2025, we introduced the Sapient Hierarchical Reasoning Model (HRM), a hierarchical, brain-inspired model that achieves deep reasoning with minimal data. With just 27 million parameters and approximately 1,000 training examples, without pre-training, Sapient HRM achieves near-perfect accuracy on Sudoku Extreme, Maze Hard, and other high-difficulty math tasks and outperforms current models that are significantly larger on the ARC-AGI. Early pilot applications will include healthcare, robot control, and climate forecasting. Our fast-growing team includes alumni of Google DeepMind, DeepSeek, Anthropic, and xAI, alongside researchers from Tsinghua University, Peking University, UC Berkeley, the University of Cambridge, and the University of Alberta, working together to close the gap between today's language models and true general intelligence. For more information, visit Media Contact [email protected], [email protected] Media Contact Company Name: Sapient Intelligence Contact Person: Gen Li Email: Send Email Country: China Website: Source: EmailWire
Yahoo
11-06-2025
- Business
- Yahoo
Publicis deepens AI transformation services with help from Nvidia
This story was originally published on Marketing Dive. To receive daily news and insights, subscribe to our free daily Marketing Dive newsletter. Publicis Groupe unveiled a Center of Excellence (COE) in partnership with Nvidia that supports clients as they reshape their enterprises around artificial intelligence (AI), according to a press release. The initiative, spearheaded by digital transformation agency Publicis Sapient, has three core objectives: building an 'AI factory' for developing diverse client use cases; upgrading Sapient's agentic AI platform, Bodhi; and launching solutions that help clients establish the infrastructure needed to manage AI strategy. The consultative approach speaks to the large opportunity agencies see in aiding brands along their journey with AI adoption. Publicis has pledged to spend hundreds of millions on AI as it tries to evolve into an 'intelligent system.' The launch of Publicis' Center of Excellence marks another step in the ad-holding group's yearslong effort to refashion itself as a tech-enabled platform versus a traditional agency network. The offering speaks to the high demand among clients for AI-driven transformation, which can be costly to realize and cumbersome to implement regarding IT software and other infrastructure. To that end, Publicis and its Sapient arm are deepening their work with chipmaker Nvidia, a leader in the current AI arms race. The idea is to blend Nvidia's tech sophistication with Sapient's transformation expertise. The COE represents a convergence of emergent AI technology with the consultative services agencies have ramped up as they look to stay competitive, not only with conventional category peers, but also consulting giants like Accenture, Deloitte and EY that have expanded their marketing divisions. Publicis plans to stand up an advanced AI factory that uses Nvidia's DGX systems, as well as a full-stack solutions service that assists with different areas of AI lifecycle management, including data ingestion, training, refinement and inference. The aim is to provide a 'a holistic approach to enterprise AI adoption,' per the announcement, bolstered by Nvidia's AI Enterprise suite. At the same time, the partnership will soup up Bodhi, Sapient's existing platform for creating AI agents. The goal is to produce virtual agents that can 'reason, plan and act independently.' 'By leveraging NVIDIA's advanced AI infrastructure alongside our Bodhi platform and digital business transformation expertise, we're creating a new paradigm for enterprise AI adoption,' said Nigel Vaz, CEO of Sapient, in a press statement. 'This partnership enables us to guide clients seamlessly through every stage of their AI journey — transforming strategy into actionable, autonomous intelligence that drives real business outcomes.' Recommended Reading Publicis' latest reinvention: an 'intelligent system' fueled by generative AI Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Business Wire
30-04-2025
- Health
- Business Wire
Sapient Launches DynamiQ™ as Its Next-Generation Data Insights Engine for Drug Discovery and Development
SAN DIEGO--(BUSINESS WIRE)-- Sapient, a leader in multi-omics data generation for biomarker discovery and clinical insight delivery, has launched the next generation of its multi-omics and real-world data (RWD) human biology database, now branded as the DynamiQ™ Insights Engine, which increases the speed, versatility, and depth of analyses that Sapient can perform to inform drug discovery and precision drug development strategies. Multi-omic measures and real-world data have been collected and integrated for more than 56,000 samples from diverse individuals across multiple timepoints, making DynamiQ among the most deeply phenotyped datasets available for longitudinal analysis. Share DynamiQ features an expanded data lakehouse infrastructure, developed in collaboration with Sapient's partner Rancho Biosciences, that enables rapid curation of Sapient's unique datasets comprising thousands of protein, metabolite, lipid, and cytokine measures per sample alongside matched genomic and RWD including electronic health records (EHR), lab measures, and clinical outcomes. This data has been collected and integrated for more than 56,000 samples from diverse individuals across multiple timepoints, making DynamiQ among the most deeply phenotyped datasets available for longitudinal analysis. Data has been harmonized across conditions, diagnoses, and outcomes, allowing for rapid creation of cohorts in which to discover, validate, and characterize biomarkers or therapeutic targets. It also provides longitudinal views of patient journeys to identify dynamic drivers of disease and drug response and to compare disease progression and treatment-response patterns across individuals and over time. 'Disease processes aren't static, and that's why DynamiQ is our differentiator,' said Dr. Jonathan Usuka, Chief Executive Officer at Sapient. 'We're focused on the deep characterization of dynamic molecular processes which are read out by proteins and metabolites that modulate, or are modulated by, disease and exposures. We have these multi-omic measures from samples collected in the same patients at different timepoints, and when we combine those with inputs from their clinical data, we can better decipher multifactorial diseases in the context of the patient's real-world experience – how disease subtypes manifest, how lifestyle factors contribute, and how drug exposure and drug adherence impact biological mechanisms and clinical outcomes. This becomes extremely powerful to understand how a therapy will actually work in a patient, to better predict response and stratify patients that are most likely to benefit from a given therapy.' 'With DynamiQ's next-generation functionality and harmonized datasets, we increase the depth and breadth of analyses we can perform for biomarker discovery, target identification, and clinical insight delivery,' said Dr. Tao Long, Co-Founder and Head of Data Science at Sapient. 'We can readily analyze molecular interactions across omics layers, rapidly build cohorts from standardized patient data, and validate discoveries from client studies in these independent populations. Importantly, because our multi-omics datasets are nontargeted they are ideal for AI and machine learning-based analysis to uncover new subgroups, novel biomarkers, and targets.' Sapient will offer guided analyses using the DynamiQ Insight Engine as part of clients' multi-omic studies or for standalone biocomputational projects aimed at elucidating novel biomarkers, drug targets, and diagnostics. About Sapient Sapient is a leader in multi-omics data generation and insight delivery, providing bespoke services for proteomics, metabolomics, and lipidomics that enable biopharma sponsors to go beyond the genome to accelerate precision drug development. Utilizing cutting-edge, high-throughput mass spectrometry and biocomputational frameworks, Sapient enables comprehensive biomarker-phenotype mapping across thousands of biosamples for discovery of robust protein, metabolite, and lipid biomarkers, drug targets, and clinical signatures of drug response. The company's DynamiQ™ Insights Engine — a longitudinal database of integrated multi-omics and real-world data collected from tens of thousands of samples — enables rapid drug target identification, biomarker discovery and validation, and translational insights across all stages of drug development. For more information, visit
Associated Press
30-04-2025
- Health
- Associated Press
Sapient Launches DynamiQ™ as Its Next-Generation Data Insights Engine for Drug Discovery and Development
SAN DIEGO--(BUSINESS WIRE)--Apr 30, 2025-- Sapient, a leader in multi-omics data generation for biomarker discovery and clinical insight delivery, has launched the next generation of its multi-omics and real-world data (RWD) human biology database, now branded as the DynamiQ™ Insights Engine, which increases the speed, versatility, and depth of analyses that Sapient can perform to inform drug discovery and precision drug development strategies. DynamiQ features an expanded data lakehouse infrastructure, developed in collaboration with Sapient's partner Rancho Biosciences, that enables rapid curation of Sapient's unique datasets comprising thousands of protein, metabolite, lipid, and cytokine measures per sample alongside matched genomic and RWD including electronic health records (EHR), lab measures, and clinical outcomes. This data has been collected and integrated for more than 56,000 samples from diverse individuals across multiple timepoints, making DynamiQ among the most deeply phenotyped datasets available for longitudinal analysis. Data has been harmonized across conditions, diagnoses, and outcomes, allowing for rapid creation of cohorts in which to discover, validate, and characterize biomarkers or therapeutic targets. It also provides longitudinal views of patient journeys to identify dynamic drivers of disease and drug response and to compare disease progression and treatment-response patterns across individuals and over time. 'Disease processes aren't static, and that's why DynamiQ is our differentiator,' said Dr. Jonathan Usuka, Chief Executive Officer at Sapient. 'We're focused on the deep characterization of dynamic molecular processes which are read out by proteins and metabolites that modulate, or are modulated by, disease and exposures. We have these multi-omic measures from samples collected in the same patients at different timepoints, and when we combine those with inputs from their clinical data, we can better decipher multifactorial diseases in the context of the patient's real-world experience – how disease subtypes manifest, how lifestyle factors contribute, and how drug exposure and drug adherence impact biological mechanisms and clinical outcomes. This becomes extremely powerful to understand how a therapy will actually work in a patient, to better predict response and stratify patients that are most likely to benefit from a given therapy.' 'With DynamiQ's next-generation functionality and harmonized datasets, we increase the depth and breadth of analyses we can perform for biomarker discovery, target identification, and clinical insight delivery,' said Dr. Tao Long, Co-Founder and Head of Data Science at Sapient. 'We can readily analyze molecular interactions across omics layers, rapidly build cohorts from standardized patient data, and validate discoveries from client studies in these independent populations. Importantly, because our multi-omics datasets are nontargeted they are ideal for AI and machine learning-based analysis to uncover new subgroups, novel biomarkers, and targets.' Sapient will offer guided analyses using the DynamiQ Insight Engine as part of clients' multi-omic studies or for standalone biocomputational projects aimed at elucidating novel biomarkers, drug targets, and diagnostics. About Sapient Sapient is a leader in multi-omics data generation and insight delivery, providing bespoke services for proteomics, metabolomics, and lipidomics that enable biopharma sponsors to go beyond the genome to accelerate precision drug development. Utilizing cutting-edge, high-throughput mass spectrometry and biocomputational frameworks, Sapient enables comprehensive biomarker-phenotype mapping across thousands of biosamples for discovery of robust protein, metabolite, and lipid biomarkers, drug targets, and clinical signatures of drug response. The company's DynamiQ™ Insights Engine — a longitudinal database of integrated multi-omics and real-world data collected from tens of thousands of samples — enables rapid drug target identification, biomarker discovery and validation, and translational insights across all stages of drug development. For more information, visit View source version on CONTACT: Media Contact: Carla Nieser Head of Marketing, Sapient [email protected] 858.290.7010 KEYWORD: UNITED STATES NORTH AMERICA CALIFORNIA INDUSTRY KEYWORD: DATA MANAGEMENT HEALTH TECHNOLOGY HEALTH TECHNOLOGY PHARMACEUTICAL BIOTECHNOLOGY SOURCE: Sapient Copyright Business Wire 2025. PUB: 04/30/2025 07:14 AM/DISC: 04/30/2025 07:13 AM
