Deep Principle

Start exploring Hydrogen-1.0. Download the open weights to adapt and evaluate the model for your own research tasks, or try it directly in MIRA.For training methods, detailed evaluations, and application examples, explore our technical blog. → Model weights: huggingface.co/deep-principl… → Try Hydrogen-1.0: mira.deepprinciple.com/ → Technical blog: blog.deepprinciple.com/intro… We look forward to seeing what you explore.

In our metal-complex search evaluation, Hydrogen-1.0 used feedback from previous recommendations to search for candidates with larger HOMO–LUMO gaps. A concrete example of using successive evaluation results to guide materials exploration toward a defined target.

Hydrogen-1.0 helps researchers develop procedures and specify individual operations. Lab Compile then checks equipment capabilities, prerequisites, and execution order in a virtual environment. One concrete example: flagging an instruction to aspirate liquid before a fresh pipette tip has been fitted. By checking these details virtually, researchers can identify and revise execution issues before moving to the lab.

Hydrogen-1.0 helps researchers investigate material properties, compare reaction pathways, and identify promising directions for experiments through calculations and simulations. In our polyethylene example, it builds a molecular model and writes simulation code. When a density calculation produces an abnormal result, it helps trace the issue to double-counting, correct the code, and rerun the calculation. These calculations give researchers a clearer basis for deciding what to test next.

In our evaluations, Hydrogen-1.0 achieved: - 38.3% problem-solving success on FrontierScience Research (FSR), 5.0 percentage points above GPT-6 Astra. - 93.0% accuracy on GPQA Diamond, close to Kimi K3’s 93.5%. These benchmarks examine two essential capabilities for scientific work: reasoning through unfamiliar, multi-step problems and answering challenging questions in physics, chemistry, and biology.

We just dropped Materials Property Axiom (MPA). The problem: most materials foundation models crush it on computed properties, but choke on real experimental data (boiling points, solubility, flammability, etc.). MPA fixes this with physics-aligned transfer learning: • 64M structures → 3D pre-training • Thermochemistry + electronics + solvation → mid-training alignment • Hybrid attention/additive readout → experimental fine-tuning Results: ~10% better across 40 benchmarks, up to 51% on hard OOD tasks. The rule: match the physics, not just the data. Live at sciclaw.ai 🔗 blog.deepprinciple.com/intro…

The article highlights the "Scientific Discovery Evaluation" (SDE) benchmark, led by Deep Principle. Unlike traditional evaluations, SDE does not consist of isolated, difficult questions. Instead, it distills 43 research scenarios and a total of 1,125 interrelated tasks from 8 ongoing, real-world research projects with unpublished data.

The article opens by pointing out that the scientific community has long questioned traditional question-and-answer evaluation benchmarks (such as HLE). These tests often focus on obscure or even trivial knowledge points, making it difficult to measure an AI's ability to conduct real scientific research.

Thrilled to be on the cover of JACS! @ACSPublications LLM‑EO (LLMs + evolutionary optimization) designs transition‑metal complexes, surfacing top candidates and novel ligands from a 1.37M space. pubs.acs.org/doi/10.1021/jac… From cover to frontier—Deep Principle keeps driving GenAI deeper into materials. More breakthroughs ahead. #AI4Science #MyACSCover

#AI is already revolutionizing #material_science . Haojun Jia, founder and CEO of Deep Principle, explains how leveraging both domain expertise and data-driven insights will be key to AI-enabled platforms transforming materials #science #AMNC25 weforum.org/stories/2025/06/…

🎉 We’re thrilled to announce that Deep Principle has been selected as a 2025 World Economic Forum Technology Pioneer! 🌍 Each year, the WEF recognizes early-stage companies poised to make a significant impact on business and society. We’re proud to be recognized alongside global innovators driving sustainable, intelligent, and accelerated growth. 🧪 At Deep Principle, we are redefining the future of chemistry and materials innovation through cutting-edge AI and first-principles computing—enabling smarter scientific discovery at unprecedented speed and scale. We’re excited to contribute to the Forum’s global initiatives and conversations, and look forward to seeing our co-founders @HaojunJia1 and @chenru_duan represent us at key WEF events throughout the year. 🔗 #TechPioneers25 #WEF #AIforScience #DeepTech #MaterialsInnovation

Deep Principle Unveils React-OT: Fast, Deterministic Transition State Prediction We’re excited to introduce React-OT, our groundbreaking AI solution for deterministic transition state (TS) generation—now published as the cover article in Nature Machine Intelligence. Engineered for chemists and materials scientists, React-OT tackles the decades-old challenge of efficiently capturing elusive TS structures, which are pivotal to decoding reaction mechanisms and designing next-generation catalysts. The Challenge: TS Search in the Age of Complex Reactions Transition states are the crucial turning points in chemical pathways, serving as the most informative snapshots for understanding reactivity and mechanism. Yet traditional methods—powered by quantum chemistry—are prohibitively slow, demanding extensive computational resources. Machine learning offered hope, yet existing approaches like diffusion models remained stochastic, cumbersome, and inconsistent in accuracy. We needed a deterministic, scalable solution to unlock the full potential of high-throughput reaction exploration. Three Key Advantages: ✅ Blazing Speed: Generates high-accuracy TS structures in 0.4 seconds/reaction—1,000x faster than traditional computational approaches. ✅ Unmatched Precision: Achieves 0.053 Å median structural error and 1.06 kcal mol⁻¹ median barrier height error; pretraining on 950k semi-empirical quantum chemistry computed reactions furthur reduces errors by up to 25%. ✅ Cost-Efficient Workflows: Compatible with low-cost GFN2-xTB geometries, reducing DFT reliance. Paired with uncertainty control, it enables accurate exploration of chemical reaction networks. Why It Matters 💡 Accelerate Discovery: Map complex reaction networks without bottlenecks from costly TS searches. 💡 Drive Innovation: Design better catalysts and unravel unknown mechanisms with confidence in structural and energetic accuracy. 💡 Scale Seamlessly: Integrate into high-throughput pipelines to prioritize expensive DFT only when needed. Join the Future of Chemistry and Material Discover how React-OT is revolutionizing TS prediction for drug discovery, energy research, and materials science. Read the full cover story here and explore how Deep Principle’s AI tools are accelerating scientific breakthroughs. You can also experience the app to explore reaction pathways, visualize transition states, and run predictions in real time. #AIforScience #ChemistryBreakthrough #MachineLearning #CatalystDesign #DeepPrinciple @DeepPrincipleAI | deepprinciple.com

Revolutionizing Chemistry with AI: A New Era of Innovation🚀 At Deep Principle Inc., we are at the forefront of transforming how chemistry and materials science are approached in the 21st century. By integrating cutting-edge AI models, Quantum Chemistry, and High-Throughput Experimental techniques, we’re reshaping the entire R&D process.🧪 Our approach moves beyond traditional trial-and-error methods, embracing a data-driven, AI-powered workflow that accelerates the discovery and development of chemical materials. The result? A quantum leap in efficiency, precision, and innovation across the scientific community.🔬 As we continue to lead the way in AI for Science, we invite you to join us on this transformative journey to reshape the future of chemistry and materials science. Together, we can unlock new possibilities and accelerate solutions for a more sustainable, innovative world！🌍 #AI #Innovation #Chemistry #MaterialsScience #Sustainability #AIforScience #FutureOfChemistry #DeepPrinciple

Introducing Deep Principle’s Visionary Team 🌍 At Deep Principle, we believe that the convergence of chemistry, engineering, and AI holds the key to solving some of the world’s most pressing challenges. Over the past few months, we’ve assembled an exceptional team of experts dedicated to pushing the boundaries of what’s possible in materials science and AI. Our inaugural team conference was a moment of reflection and inspiration. Bringing together such a diverse group of thinkers – from seasoned chemists to cutting-edge AI researchers – was a reminder of our shared purpose: to accelerate innovation and deliver impactful, sustainable solutions for the industries that shape our future. A special thanks to our talented team for your unwavering commitment, boundless energy, and belief in our mission. @HaojunJia1, @chenru_duan, @ZhaoQiyuan, @Artistypl and etc. Together, we are reimagining the future of materials discovery, catalyzing the shift towards more sustainable and efficient solutions across a wide range of industries. Stay tuned as we continue to push the limits of AI and materials science, solving complex problems and shaping the future. 🌱✨ #AI #MaterialsScience #Innovation #Sustainability #Teamwork #FutureOfChemistry #ClimateTech #DeepPrinciple