معرفی
Marta Halina is a University Associate Professor in the Philosophy of Cognitive Science at the University of Cambridge, affiliated with the Department of History and Philosophy of Science. She serves as a Senior Research Fellow at the Leverhulme Centre for the Future of Intelligence and is a Fellow of Selwyn College. Her academic journey began with a PhD in Philosophy and Science Studies from the University of California, San Diego in 2013, followed by a McDonnell Postdoctoral Fellowship in the Philosophy-Neuroscience-Psychology Program at Washington University in St. Louis before joining Cambridge in 2014.
Halina's educational background includes a PhD from UC San Diego (2013) and postdoctoral training at Washington University in St. Louis. Her academic trajectory reflects a strong interdisciplinary foundation bridging philosophy, cognitive science, and neuroscience.
Her research focuses on nonhuman animal cognition, mechanistic explanation, and artificial intelligence, with particular emphasis on comparative cognition and the philosophical foundations of cognitive science. Halina investigates how researchers design studies to address complex questions about animal minds, arguing that current methods in comparative cognition often face challenges with hypothesis underdetermination by empirical evidence. She advocates for additional behavioral constraints on theorizing, known as 'signature testing,' while emphasizing the need to incorporate neuroscience and biology more substantially into animal cognition research. Her work on major transitions in cognitive evolution proposes treating the evolution of cognition as a series of major evolutionary transitions to better comprehend cognitive complexity across species.
Analysis of Halina's recent publications reveals a clear trajectory toward computational comparative cognition. Her work increasingly integrates AI and machine learning techniques with traditional comparative cognition approaches, exemplified by her development of the Animal-AI Testbed. This platform allows for direct comparison between AI systems, humans, and animals on cognitive tasks, revealing that while AI and children perform similarly on basic navigational tasks, children outperform AI on more complex cognitive tests requiring object permanence. Her research demonstrates how computational modeling can generate novel hypotheses about animal behavior that generate precise, testable predictions beyond what traditional experimental methods alone can achieve.
- McDonnell Postdoctoral Fellowship
Halina directs research initiatives at the Leverhulme Centre for the Future of Intelligence, particularly focusing on the intersection of AI and animal cognition. Her work on the Animal-AI Environment has received significant funding and collaborative support, enabling interdisciplinary research that bridges computer science, cognitive science, and biology. She actively collaborates with researchers across multiple institutions to develop computational frameworks for understanding nonhuman animal cognition.
Halina leads significant research initiatives through the Leverhulme Centre for the Future of Intelligence, where she develops the Animal-AI Environment—a research platform for conducting cognitive experiments with artificial agents, humans, and nonhuman animals in directly comparable, ecologically valid contexts. This environment facilitates interdisciplinary collaboration between computer scientists, engineers, biologists, and cognitive scientists, reducing the 'language barrier' between these fields and enabling cross-pollination of ideas and methodologies.
