Advancing the foundations of language understanding through neural-symbolic approaches, computational linguistics, and hybrid AI systems.
Basement AI is an independent research lab bridging theoretical linguistics and machine learning. We develop foundational models that combine the interpretability of symbolic systems with the power of neural networks to advance language understanding.
Our research spans foundational questions in computational linguistics, hybrid AI architectures, and the development of robust, interpretable language models.
Developing neural-symbolic architectures that combine the learning capabilities of neural networks with the interpretability and reasoning power of symbolic systems for robust language understanding.
Bridging linguistic theory with machine learning through structured approaches to language modeling, exploring how grammatical formalisms can inform more interpretable AI systems.
Investigating how language models behave under different conditions—from data corruption to domain shifts—to develop more reliable and generalizable NLP systems for real-world applications.
Selected publications developing foundational approaches to language understanding, model reliability, and reasoning.
Want to connect about research, ventures, or collaborations? We'd love to hear from you.