Research Intern: Meta-Cognitive Agentic AI Systems

Job Number: P25INT-49 Honda Research Institute USA (HRI-US) is seeking a self-motivated research intern to join our Cooperative Cognition Team in advancing the next generation of adaptive AI systems that can thrive in complex, human-AI collaborative environments. As an intern, you will help design and implement AI agents with metacognitive capabilities, focusing on areas such as multi-agent systems for long-horizon tasks, test-time scaling, and rapid adaptation to individual preferences. This work will contribute to developing robust, self-adapting and self-correcting AI partners that can quickly learn new tasks, adjust to dynamic teams, and cooperate effectively with humans. This includes exploring how internal model representations—such as hidden states, neural activations, and latent structures—can be analyzed and leveraged for self-monitoring, uncertainty estimation, and adaptive control in meta-cognitive agent architectures. San Jose, CA **Key Responsibilities** * Contribute to the development of metacognitive algorithms (assessment and intervention strategies) that enhance the efficiency, adaptability, and reliability of agentic AI systems. * Develop methods to analyze, probe, and manipulate neural activations and internal representations of foundation models (e.g., hidden states and latent features), including activation-level and circuit-level analysis, to enable meta-cognitive capabilities such as self-assessment, uncertainty awareness, and adaptive intervention. * Test-time scaling of multimodal (vision–language–action) models. * Improve the robustness and generalization of LLMs in long-horizon, multi-agent tasks. * Design and curate a benchmark suite for evaluating multimodal agentic systems in collaborative settings. **Minimum Qualifications** * Currently enrolled as a Ph.D. student in Computer Science, Machine Learning, or a related field at a reputed university (exceptional M.S. candidates with a minimum of 1 year of research experience may also be considered. * Strong familiarity with agentic AI systems, modern foundation models, and their internal representations (e.g., hidden states, neural activations, embeddings, and latent structures). * Experience in open-source deep learning frameworks. **Bonus Qualifications** * Experience with representation learning, mechanistic interpretability, or analysis of neural activations in foundation models. Familiarity with techniques for probing, interpreting, or steering internal model behavior (e.g., activation-level analysis, feature attribution, or circuit-level analysis). * **Years of Work Experience Required** 0 **Desired Start Date** 8/31/2026 **Internship Duration** 3 Months **Position Keywords** Agentic AI, Meta-Cognitive AI **Warning:**You do not have the permission to access the upload fields on this form. Contact the form owner or portal administrator to request the access.