AI Benchmark & Datasets Engineer/Researcher Internship

About PathwayAt Pathway, we are pioneers in AI technology, developing the first post-transformer frontier model that addresses the fundamental memory challenges faced by traditional AI systems. Unlike conventional transformers that reset each time, our innovative architecture allows for genuine continuous learning, extensive contextual reasoning, and real-ti…

Location: Remote or Palo Alto, CADuration: 12–16 weeks (flexible)Compensation: Paid, competitiveStart: RollingAbout PalonaAt Palona, we are dedicated to creating real-world AI systems that operate continuously in production environments. Our focus is on developing AI agents that can perceive, reason, remember, and act in physical spaces, starting with restaurants as our initial domain due to its complexity and high signal density.We thrive on research that stands the test of reality, addressing challenges such as partial observability, delayed effects, noisy signals, non-stationarity, and long-term outcomes.Research ScopeThis internship is designed for PhD students eager to tackle applied research challenges linked to deployed systems. You will explore questions stemming from live AI agents functioning in the real world, where ideal assumptions may fail, and understanding system behavior over time is essential.Required Research Background (PhD Level)We seek candidates with extensive research experience in at least one primary area, alongside a working knowledge of related fields.Primary Research Areas (at least one required)1. Sequential Decision MakingReinforcement learning, planning, or controlPOMDPs or decision-making under partial observabilityCredit assignment with delayed and sparse rewardsLong-horizon optimizationRelevant indicators:Publications in RL, planning, or control venuesExperience in implementing and evaluating decision-making agents2. World Modeling and State RepresentationLatent state models for dynamic environmentsTemporal abstraction and hierarchical representationsPersistent memory or state trackingModeling environments that evolve over timeResearch in state-space models, memory-augmented models, or temporal representations3. Reasoning Under Uncertainty and CausalityBelief state estimationUncertainty modeling in dynamic systems with incomplete or noisy informationResearch in probabilistic modeling, causal inference, or dynamic systems4. Multimodal Learning in Real EnvironmentsVision-language modelsLearning from asynchronous, noisy, or partially missing modalitiesSensor fusion or multimodal representation learningPublications or projects involving multimodal modelsExperience working with real-world (not solely synthetic) dataWhat You Will Work OnYour projects will be tailored to your expertise and may encompass:Designing AI systems that effectively navigate complex environments

About UsAt Odyssey, we are at the forefront of artificial intelligence research, developing cutting-edge general-purpose world models. These innovative multimodal intelligence frameworks are set to transform consumer, enterprise, and intelligence applications. Our flagship model, Odyssey-2 Pro, exemplifies our commitment to pioneering advancements in AI.OpportunityWe are seeking passionate research interns for 2026 to collaborate with our expert machine learning research teams. This internship is designed for PhD students eager to expand the horizons of visual AI by developing and executing a research project focused on enhancing general-purpose world models. You will have access to Odyssey's extensive research and engineering resources, data, and computational capabilities, making a tangible impact on our publicly available models.Candidate ProfileThe ideal candidate is currently pursuing a PhD or a similar advanced degree and has relevant research experience to share. You should be available to work from our offices in Palo Alto or London, with flexibility to conduct research at any time throughout the year.Your RoleDuring this internship, you will undertake a 3-6 month research project aimed at producing publishable results in the field of general-purpose world models. We welcome a diverse range of research topics that contribute to advancing visual AI. This project will enhance your academic portfolio and will be supported by mentorship from our research teams and technical infrastructure to ensure your success.Application ProcessTo apply, please submit a concise research proposal outlining the project you wish to pursue, limited to one page. This proposal will help us shape the scope of potential projects in collaboration with selected candidates and the Odyssey research team.

At genbio, a cutting-edge start-up based in Silicon Valley, we unite visionary scientists, engineers, and entrepreneurs who are passionate about reshaping biology and medicine with the innovative potential of generative AI. Our team is comprised of leading experts and trailblazers in AI and biological sciences, continually striving to push the frontiers of what's achievable. We are the dreamers who are re-envisioning the future of biology and medicine.Our mission is to comprehensively decode biological processes, paving the way for transformative health solutions. As pioneers in pan-modal Large Biological Models (LBM), we are at the forefront of a new era in biomedicine, where our LBM training is catalyzing groundbreaking advancements and reshaping healthcare. With a robust R&D team and a leadership role in LLMs and generative AI, we are well-positioned to make a significant global impact. Join us on this exciting journey as we redefine the future of biology and medicine through the transformative power of Generative AI.

At Rhoda AI, we are pioneering the development of a comprehensive foundation for the next generation of humanoid robots. Our approach integrates high-performance, software-defined hardware with advanced models and world models that facilitate their operation. Our robots are engineered to function as generalists, adept at navigating complex, real-world environments and managing previously unseen scenarios. We are at the forefront of large-scale learning, robotics, and systems research, supported by a diverse team of experts from prestigious institutions including Stanford, Berkeley, and Harvard. With over $400 million raised, we are committed to investing in the R&D, hardware innovation, and manufacturing scale-up necessary to bring our vision to life.We invite applications for the position of Research Engineer, where you will collaborate closely with our research team on comprehensive model development. This hands-on role encompasses the entire stack: data management, infrastructure, model training, and deployment. You will play a critical role in transforming research concepts into scalable, operational systems, including the learning and application of world models for planning, prediction, and control.Key ResponsibilitiesDesign and develop foundational and world models for extensive robotic learning.Establish and manage data pipelines encompassing collection, curation, filtering, and augmentation for multimodal robotic data (vision, proprioception, actions, language, video).Engage in pre-training and post-training processes, including fine-tuning, alignment, and evaluation of large models and world models.Implement and experiment with various model architectures.Create training and evaluation frameworks for world models, focusing on rollout quality, long-horizon predictions, and downstream task performance.Enhance training infrastructure and workflows (distributed training, efficiency, debugging).Collaborate closely with researchers to convert ideas into resilient, scalable implementations.Assist with experiments, ablations, and real-world deployments on robotic systems.QualificationsProficiency in software engineering combined with a research-driven mindset.Demonstrated experience in implementing ML models end-to-end, beyond merely executing existing code.Comprehensive understanding of the entire ML pipeline: data → pre-training → post-training → evaluation → deployment.Strong foundation in deep learning frameworks and methodologies.Ability to work collaboratively in a fast-paced, innovative environment.

AI Research Engineer specializing in Reinforcement Learning | AI & RoboticsLocation: Palo Alto, CA (on-site)About 1XAt 1X, we are pioneering the development of humanoid robots that collaborate with humans to address labor shortages and foster abundance in various industries.The RoleAs an AI Research Engineer with a focus on Reinforcement Learning, you will play a vital role in enhancing NEO's capabilities through advanced RL algorithms. This position involves working in both simulated and real-world environments to create robust behaviors and implement them within home settings. Your contributions will be crucial in increasing the safety, efficiency, and versatility of our robotic systems.

AI Research Engineer, World ModelsPalo Alto, CA (On-site)About 1XAt 1X, we are pioneering the development of humanoid robots designed to collaborate with humans, addressing labor shortages and fostering abundance across various industries.The RoleAs an AI Research Engineer specializing in world models, you will create expansive multi-modal generative models that project future sensor inputs and robotic actions derived from historical data. These foundational models empower robots to comprehend and navigate complex real-world environments. Your responsibilities will span data engineering, model architecture, and deployment, with the goal of enhancing robot autonomy. This position merges innovative research with pragmatic product development, challenging the boundaries of robotic intelligence.

About PathwayPathway is revolutionizing artificial intelligence with the introduction of the world’s first post-transformer model that mimics human thought processes. Our innovative architecture surpasses traditional Transformer models, providing enterprises with unparalleled transparency into model operations. By integrating this foundational model with the fastest data processing engine available, Pathway empowers organizations to transcend mere incremental optimization and achieve genuinely contextualized, experience-driven intelligence. Trusted by prestigious clients including NATO, La Poste, and Formula 1 racing teams, we are at the forefront of AI advancements.Led by visionary CEO Zuzanna Stamirowska, a complexity scientist, our team includes AI trailblazers such as CTO Jan Chorowski, who pioneered the application of Attention in speech and collaborated with Nobel laureate Geoff Hinton at Google Brain, and CSO Adrian Kosowski, a distinguished computer scientist and quantum physicist who earned his PhD at just 20 years old.Supported by prominent investors and advisors like Lukasz Kaiser, co-author of the Transformer architecture (the “T” in ChatGPT) and a key researcher in OpenAI's reasoning models, Pathway is headquartered in Palo Alto, California.The OpportunityWe are on the lookout for passionate Machine Learning/AI Software Engineering interns with a solid foundation in machine learning model research.Your ResponsibilitiesAssist in training Large Language Models (LLMs)Conduct benchmarking of LLMsPrepare and evaluate training datasetsCollaborate with the core Pathway Research TeamYour contributions will significantly impact the advancement of the AI landscape.

Position OverviewJoin our innovative team at Nace.ai as we push the boundaries of artificial intelligence through cutting-edge research in large language models (LLMs) and vision-language models (VLMs). We are in search of a talented AI Research Engineer with a strong focus on adaptive learning methodologies, including meta-learning and hypernetworks. Your role will encompass the design and implementation of advanced architectures for dynamic model adaptation, enhancing model reasoning capabilities, and effectively sharing insights with both research and engineering teams.Essential Qualifications:Demonstrated experience with LLMs or VLMs in both research and production environments.Strong foundational knowledge in Natural Language Processing, Machine Learning, or related fields, particularly in language model development.A proven history of tackling complex challenges in language understanding and generation, employing rigorous quantitative methods.Exceptional communication skills for conveying research findings to varied technical audiences.Proficiency in Python and familiarity with deep learning frameworks such as PyTorch, JAX, or TensorFlow, alongside experience in distributed training and model optimization.Desirable Qualifications:PhD in Computer Science, Computational Linguistics, or a closely related discipline with an emphasis on language models and adaptive learning frameworks.Substantial research and engineering background with LLMs/VLMs, particularly in meta-learning or parameter-efficient adaptation, supported by grants, fellowships, patents, or contributions to open-source initiatives.First-author publications in recognized peer-reviewed conferences (ACL, EMNLP, NeurIPS, ICML, ICLR) or journals focusing on language models, meta-learning, hypernetworks, or adaptive AI.Preferred Technical Expertise:In-depth research knowledge in LLM reasoning, hypernetworks, multi-task learning, meta-learning, and the design of innovative LLM adaptation techniques, including online continual learning.

At code-metal, we are on a mission to revolutionize hardware deployment by matching the rapid pace of software development. Our innovative work focuses on automatic code transpilation and optimization for diverse hardware applications.As an AI Research Engineer, you will work alongside a dynamic team of researchers, tackling groundbreaking projects in generative AI and reinforcement learning.Core Responsibilities:Independently design, execute, and analyze complex experiments.Contribute to the development of core models and frameworks.Generate high-quality datasets, both real-world and synthetic.Conduct literature reviews and implement cutting-edge techniques from research papers.Engage in the publication process and present findings at conferences and workshops.Research Areas of Interest:Our current and near-term research focuses include:Contrastive representation learningSteerability and guided decodingTractable probability modelsCode-specific architecturesLLM fine-tuning, post-training, RLHF

AI Research Engineer, Scaling | InfrastructureLocation: Palo Alto, CA (on-site)At 1X, we are pioneering the development of humanoid robots designed to collaborate with humans, addressing labor shortages and fostering abundance across various sectors.The Role: As an AI Research Engineer specializing in Scaling, you will be responsible for architecting and implementing robust infrastructure that facilitates large-scale training, evaluation, and deployment for our fleet of robots. Your contributions will be essential in transitioning experimental systems into production-ready platforms, optimized for throughput, latency, and overall performance in both datacenter and edge environments. This role will significantly impact the efficiency of learning and inference processes, directly influencing the capabilities of our general-purpose humanoid robots.

About UsOdyssey is an innovative AI laboratory at the forefront of developing general-purpose world models. These advanced multimodal intelligence systems are set to revolutionize consumer, enterprise, and intelligence applications. With models like Odyssey-2 Pro, we are pioneering the next significant leap in AI technology.Position OverviewWe are in search of passionate engineers who excel in the art of building robust systems. You should possess the ability to write elegant, scalable machine learning code, with a strong emphasis on performance and an understanding of the underlying research. You are comfortable navigating the realms of modeling and systems, boldly tackling complex technical challenges while taking pride in constructing the infrastructure and tools that enable groundbreaking advancements.Your ResponsibilitiesDevelop and scale the training and inference systems that drive Odyssey’s general-purpose world models, encompassing large-scale distributed pipelines and real-time optimization.Collaborate closely with researchers to prototype novel architectures, enhance model performance, and transition concepts from research to production.Create high-performance data and computation systems for video generation and control, facilitating rapid iteration and effective resource utilization.Design tools, metrics, and visualizations that provide insights into model behavior and evolution.Work hand-in-hand with product engineers to incorporate Odyssey’s models into real-time, interactive user experiences that exemplify new general-purpose world models.Embrace a fast-paced iterative approach. As part of a tightly-knit team, your experiments will evolve into demos and ultimately into products.Contribute to shaping Odyssey’s engineering culture, which is pragmatic, research-oriented, and always focused on what is possible next.Your ProfileA staff-level or senior engineer experienced in large-scale machine learning systems, distributed training, performance optimization, or model deployment.Hands-on and technically adept: you thrive on writing code, optimizing processes, and enhancing system efficiency.Proven experience with data structures, algorithms, and coding practices that lead to high-performance outputs.

At Simular, we are at the forefront of AI research, pushing the boundaries of what is possible in machine learning and artificial intelligence. We are seeking a passionate and driven PhD Research Intern to join our innovative team. This position may be based in any of our listed locations, with priority determined according to the order of listing.Your Role:Work closely with our talented research scientists to enhance methodologies in the following areas:Planning and Reinforcement Learning (RL) for computer applications, including behavioral cloning and RL on model weights.Multimodal grounding, focusing on vision-only models and hybrid methods incorporating large models.Reward and Judge Modeling, encompassing error analysis and human evaluation.Understanding user intent, particularly in modeling vague queries and preference learning.Assist in dataset development, conduct experiments, and benchmark results.Investigate innovative approaches to support Simular's long-term technical roadmap.Document and share findings through detailed internal reports and academic-style writing.

About Mistral AIAt Mistral AI, we harness the transformative power of artificial intelligence to streamline tasks, save valuable time, and foster enhanced creativity and learning. Our innovative technology is crafted to effortlessly integrate into everyday work environments.We are committed to democratizing AI by offering high-performance, optimized, open-source models, products, and solutions. Our extensive AI platform caters to both enterprise and individual needs, featuring products like Le Chat, La Plateforme, Mistral Code, and Mistral Compute—creating cutting-edge intelligence accessible to all users.As a vibrant and collaborative team, we are driven by our passion for AI and its potential to revolutionize society. Our diverse workforce excels in competitive settings and is dedicated to fostering innovation. With teams distributed across France, the USA, the UK, Germany, and Singapore, we pride ourselves on our creativity, humility, and team spirit.Join us in shaping the future of AI at a pioneering company. Together, we can create a lasting impact. Discover more about our culture at https://mistral.ai/careers.Role OverviewAbout the Research Engineering TeamThe Research Engineering team operates across Platform (shared infrastructure & clean coding practices) and Embedded (integrated within research squads). Our engineers have the flexibility to navigate the research↔production spectrum as their interests and needs evolve.As a Machine Learning Research Engineer, you will be responsible for building and optimizing large-scale learning systems that underpin our open-weight models. Collaborating closely with Research Scientists, you may join either:- Platform RE Team: Focus on enhancing our shared training frameworks, data pipelines, and tools utilized across all teams; or- Embedded RE Team: Become part of a research squad (Alignment, Pre-training, Multimodal, etc.) to turn innovative ideas into scalable, repeatable code.Key Responsibilities• Support researchers by managing the complex aspects of large-scale ML pipelines and developing robust tools.• Bridge cutting-edge research with production: integrate checkpoints, optimize evaluations, and create accessible APIs.• Conduct experiments utilizing the latest deep-learning techniques (sparsification on 70B+ models, distributed training across thousands of GPUs).• Design, implement, and benchmark ML algorithms; produce clear and efficient code in Python.• Deliver prototypes that evolve into production-grade components for Le Chat and our enterprise API.

AI Research Engineer, Data InfrastructureLocation: Palo Alto, CA (on-site)About 1XAt 1X, we are at the forefront of innovation, developing humanoid robots designed to collaborate with humans, effectively addressing labor shortages while fostering abundance across industries.The RoleAs an AI Research Engineer specializing in Data Infrastructure, you will play a pivotal role in designing and implementing a comprehensive data engine to efficiently manage the vast data generated by our humanoid robot fleet. Your contributions will ensure that this data is readily accessible for querying and training, supporting the development of high-quality data pipelines that facilitate effective model training, large-scale data annotation, and seamless integration across robotic, on-premise, and cloud-based systems.

About VoltaiAt Voltai, we are pioneering the development of world models and agents capable of learning, evaluating, planning, experimenting, and interacting with the physical realm. Our journey begins with a focus on hardware, specifically in electronics systems and semiconductors, where we harness AI to design and innovate beyond human cognitive capabilities.About the TeamOur team boasts extraordinary talent, including esteemed former Stanford professors, SAIL researchers, and medalists from prestigious competitions like IPhO and IOI. We are supported by top-tier investors from Silicon Valley and industry leaders, including CEOs and Presidents from Google, AMD, Broadcom, and Marvell.About the RoleAs a Research Engineer specializing in CUDA Kernel engineering, you will design, integrate, and optimize cutting-edge CUDA kernels that drive AI models, facilitating rapid advancements in semiconductor design and verification. Your contributions will empower extensive model training, inference, and reinforcement learning systems capable of reasoning about circuit layouts, generating and validating RTL, and optimizing chip architectures, all while efficiently utilizing thousands of GPUs.You will create tools, performance benchmarks, and integration layers that maximize GPU utilization for compute-intensive workloads in AI-driven hardware design. Collaborating closely with fellow researchers and engineers, you will help position Voltai as the foremost organization in AI and semiconductor research. Furthermore, your kernels and tools will be released as valuable contributions to the open-source AI and HPC ecosystems.You might excel in this position if you possess experience in:Writing and optimizing CUDA kernels for large-scale AI applications (e.g., attention mechanisms, routing, graph-based operations, and physics-inspired operators).Profiling and enhancing GPU performance for specialized compute or memory-bound workloads.Integrating custom kernels into state-of-the-art training and inference frameworks (including PyTorch, Megatron, vLLM, and TorchTitan).Engaging with the latest NVIDIA hardware and software frameworks (Hopper, Blackwell, NVLink, NCCL, Triton).Creating GPU-accelerated primitives for graph reasoning, symbolic computation, or hardware simulation tasks.

Join Our TeamAt Parallel, we are at the forefront of web infrastructure innovation, empowering businesses in various sectors—sales, marketing, insurance, and technology—to develop sophisticated AI agents equipped with robust programmatic access to the internet.Having secured $130 million in funding from prestigious investors such as Kleiner Perkins, Index Ventures, Spark Capital, Khosla Ventures, First Round, and Terrain, we are building a premier team of engineers, designers, marketers, sales professionals, researchers, and operational specialists to fulfill our ambitious vision.Your ProfileWe are looking for a researcher who embodies an engineering mindset, or an engineer who approaches problems with curiosity typical of researchers. You may have experience with information retrieval systems, embedding models, or neural ranking at scale, or possess a deep interest in the challenges of training models to comprehend and navigate billions of web pages. You will excel in the intersection of theory and practical application, devising elegant solutions that perform efficiently on real-world infrastructure. You'll be equally comfortable reading the latest papers from SIGIR and RecSys as you are troubleshooting distributed training pipelines.Position OverviewIn this role, you will design and train models that drive Parallel's APIs—the intelligent framework that enables AI agents to extract precise information from the open web. This involves addressing complex research challenges that most labs only encounter at scale: How can we create embedding models that accurately represent semantic intent across various query types? How do we achieve a balance between model expressiveness and sub-second retrieval times? How can we ensure our index remains up-to-date with the constantly evolving web, without the need for complete rebuilds?Unlike conventional search engines tailored for human queries, you will be developing solutions for AI agents that generate intricate, multi-hop queries, requiring structured, programmatic responses. This is information retrieval redefined for the era of large language models, merging traditional information retrieval methods with cutting-edge deep learning, applied at a scale that necessitates innovative solutions.Working EnvironmentOur team collaborates fully in-person at our headquarters in Palo Alto and our San Francisco office. We pride ourselves on being a flat, talent-rich organization committed to tackling both technical and creative challenges.We are eager to welcome individuals who share our enthusiasm for leveraging science, creativity, and consistency to address large, complex problems with significant impacts. Here are our core values:Customer Impact Ownership: We take responsibility for delivering tangible results for our clients.

About VoltaiAt Voltai, we are pioneering advancements in artificial intelligence by developing sophisticated world models and agents that learn, evaluate, plan, and interact with the physical environment. Our primary focus lies in enhancing hardware capabilities, particularly in electronics systems and semiconductors, where AI can surpass traditional human cognitive limitations in design and creation.About the TeamOur team comprises exceptional talent, including former Stanford professors, acclaimed SAIL researchers, Olympiad medalists, and industry leaders from renowned companies such as Google, AMD, and Broadcom. We are supported by top investors from Silicon Valley and have a diverse group of experts, including former U.S. government officials, committed to driving innovation in AI and hardware design.Role OverviewAs a Post-Training Research Engineer, you will focus on post-training cutting-edge models to autonomously execute intricate tasks within the semiconductor design and verification pipeline. The models you help develop will optimize chip architectures, refine RTL code, conduct simulations, identify verification gaps, and iteratively enhance designs to expedite semiconductor innovation. You will work alongside leading experts in hardware design and verification, crafting comprehensive reinforcement learning environments that encapsulate the complexities of chip design workflows. Your contributions will involve developing structured reward functions, scaling strategies, and evaluation frameworks aimed at enhancing model reliability, efficiency, and creativity in semiconductor reasoning.Ideal Candidate ProfileYou may excel in this role if you possess experience in:Creating and scaling reinforcement learning environments for large language models or multimodal agents.Building high-quality evaluation datasets and benchmarks for complex reasoning or design challenges.Collaborating closely with domain experts in hardware and verification to establish evaluation metrics, constraints, and simulation conditions.Designing reward functions and feedback pipelines that ensure a balance between correctness, performance, and design efficiency.Conducting large-scale reinforcement learning fine-tuning or post-training experiments on frontier models.

About VoltaiAt Voltai, we are pioneering the development of advanced world models and intelligent agents that learn, evaluate, and interact with the physical environment. Our initial focus lies in understanding and enhancing hardware, particularly in electronics systems and semiconductors, where AI surpasses human cognitive capabilities in design and innovation.About the TeamOur team is comprised of elite professionals backed by top investors in Silicon Valley, Stanford University, and industry leaders including CEOs and Presidents from Google, AMD, Broadcom, and Marvell. We bring together former Stanford professors, SAIL researchers, Olympiad medalists, and high-ranking officials from the U.S. government, all working collaboratively towards groundbreaking advancements.Mid-Level Training OpportunityIn this role, you will play a crucial part in training cutting-edge models to become experts in semiconductor design and verification, laying the groundwork for reinforcement learning and automated chip development. You will innovate methods for generating and curating synthetic design data, executing model distillation, and facilitating scalable continual learning. Collaboration will be key, as you will partner with hardware engineers, reinforcement learning researchers, and verification specialists to optimize design data quality and enhance model performance. You will also work alongside compute engineers to efficiently scale training across thousands of GPUs and RL environments, developing high-performance tools to analyze how data and simulations influence model-driven design intelligence.Ideal Candidates Will Have Experience In:Training large language models or foundation models on semiconductor design and verification datasets (e.g., RTL, netlists, PDKs, simulation logs)Modeling design scaling laws and optimizing compute budgets for chip-design-specific tasksGenerating extensive synthetic design data (e.g., RTL variations, testbenches, verification traces)Developing evaluations that correlate with downstream design metrics (e.g., timing closure, power efficiency, area, verification coverage)

Situated in the heart of Silicon Valley, Genbio is an innovative start-up uniting a diverse team of forward-thinking scientists, engineers, and entrepreneurs. Our mission is to revolutionize biology and medicine through the transformative capabilities of Generative AI. Our collective expertise includes leading minds in AI and Biological Science, dedicated to redefining the boundaries of possibility. As pioneers in the realm of pan-modal Large Biological Models (LBM), we are setting a new standard in biomedicine and healthcare. With our groundbreaking foundation model training, we aim to unlock life-changing solutions and insights into biology. With our main office located in Silicon Valley and an additional office in Paris, we are on a path to making a significant impact globally. Join us as we work to reshape the future of biology and medicine with the power of Generative AI.