At the core of BehaVERT is a transformative idea: animal movements contain latent structures analogous to human language. Rather than viewing animal behavior as isolated actions, the researchers approached it as sequences imbued with contextual meaning. By transforming skeletal keypoints—such as the movements of a mouse’s nose, ears, spine, limbs, and tail—into discrete tokens, BehaVERT treats behavioral sequences like sentences formed from words, enabling deep semantic analysis.

BehaVERT employs a BERT-based transformer architecture, a state-of-the-art deep learning model originally developed for natural language processing tasks. Transformers excel at capturing relationships across sequences by self-attention mechanisms, which allow the model to weigh the importance of different tokens contextually. This architecture empowers BehaVERT to not only classify behaviors frame by frame but also to comprehend overarching behavioral states spanning across entire sequences, thereby mimicking the way language models infer meaning beyond individual words.

To train this model, the KAIST team annotated skeletal movements from video footage using a web-based tool, converting them into high-dimensional tokens—each frame yielding 768-dimensional representations. Leveraging a self-supervised learning framework, BehaVERT was able to learn patterns directly from behavioral data without requiring extensive manual labeling. This approach reduced bias and enabled the system to uncover intrinsic behavioral semantics inherent in the data.

The robustness of BehaVERT was validated across five international benchmark datasets, demonstrating its versatility in analyzing social interactions, multi-animal dynamics, complex three-dimensional motions, and autism-related behavioral profiles. Notably, the model’s performance surpassed existing behavioral classification methods, reflecting its capacity to understand nuanced motion patterns that are often imperceptible to human observers.

One of BehaVERT’s most compelling achievements is its ability to provide interpretability. The model’s attention mechanisms can be visualized along behavioral timelines, indicating which specific motions influenced decision-making. This interpretability is crucial for bridging the gap between AI predictions and biological understanding, allowing neuroscientists to validate or discover new behavioral markers.

A striking example of BehaVERT’s interpretive power came from its analysis of an autism mouse model (Shank3B knockout mice). The AI independently highlighted oral-oral contact behaviors as key differentiators between autistic and healthy mice. This is particularly significant because it aligns with established biological findings—autism model mice exhibit social interaction deficits despite normal approaches, a subtlety that the AI rediscovered solely from raw behavioral cues without prior biological knowledge.

Further analysis showed that BehaVERT’s internal representation space intrinsically structured behavioral attributes such as mobility, attention, and social engagement into coherent clusters. This discovery suggests that animal behavior may possess an underlying semantic architecture reminiscent of linguistic grammar and syntax, shedding new light on how complex behaviors are organized and understood.

Interdisciplinary collaboration was pivotal in the development of BehaVERT. Remarkably, the lead researchers, including first author Dr. Seungjae Shin, were primarily trained in biological sciences rather than AI, exemplifying a successful fusion of domain expertise. Their dedicated self-education in transformer architectures and deep learning underpinned the creation of customized models and training approaches tailored for decoding motion semantics in rodents.

BehaVERT represents not just a new tool for behavioral classification but a platform for scientific discovery. By enabling interpretable behavioral decoding, it heralds applications in drug discovery, psychiatric disorder research, and behavioral genetics. The capacity to interpret the “language” of motion allows researchers to formulate hypotheses grounded in data-driven insights, accelerating the pace of experimental biology.

An additional technological feat demonstrated by the team is the adaptability of BehaVERT across species. A model trained on rat behavior successfully generalized to mice, highlighting the potential to develop behavioral foundation models with cross-species applications. This adaptability could revolutionize comparative behavioral studies, providing scalable analytic frameworks for diverse animal models.

KAIST’s ongoing commitment to AI-driven behavioral science is evidenced by their prior work on AVATAR, which reconstructs rodent behavior in virtual environments to facilitate detailed study. BehaVERT builds on this legacy by adding semantic understanding to motion analysis, marking a significant milestone toward holistic behavioral neuroscience.

In summary, BehaVERT is a visionary AI system that reads animal behavior as if it were a language, uncovering hidden meanings and structures within seemingly erratic motions. Its integration of skeletal tokenization, transformer architectures, and self-supervised learning delivers a potent tool bridging computational science and biology. As behavioral data continue to expand in scale and complexity, BehaVERT promises to revolutionize how we decode the fundamental language of life.

Subject of Research: Not explicitly stated

Article Title: BehaVERT: A Transformer-Based Motion Language Model for Decoding Behavioral Semantics in Mice

News Publication Date: 24-Mar-2026

Web References: http://dx.doi.org/10.1007/s11263-026-02834-y

References:
BehaVERT: A Transformer-Based Motion Language Model for Decoding Behavioral Semantics in Mice, International Journal of Computer Vision, DOI: 10.1007/s11263-026-02834-y

Image Credits: KAIST

Keywords: AI, animal behavior, transformer, BERT, motion language model, behavioral semantics, neuroscience, autism mouse model, deep learning, self-supervised learning, skeletal keypoints, behavioral analysis

The implications of such dialectical divergence extend far beyond mere communication challenges. If individual animals fail to understand or respond appropriately to the signals of others, their ability to form friendships, social bonds, or warn each other of predators or environmental hazards could be compromised. This revelation prompts a profound reevaluation of how social behaviors and traditions in nonhuman animals are maintained and transmitted across populations. More critically, it opens up the possibility that many wild animals possess culturally distinct behaviors that are not solely dictated by genetics or instinct but are acquired through social learning and tradition.

This insight into animal dialects forms part of a broader and increasingly compelling field of research that scrutinizes the cultural variability among wild species. Similar questions are being raised about chimpanzees, where some groups exhibit tool-using behaviors such as stone tool mastery while others do not. In parallel, studies of rats specializing in the consumption of pinecones suggest behavioral traditions that evolve according to local environmental conditions and social influences. These patterns underscore the importance of culture—defined broadly as socially learned behaviors that are passed between individuals—as a critical component of animal ecology and evolution.

Understanding how culture shapes animal behavior is also essential for conservation biology. As human-driven environmental changes accelerate, they threaten not just the survival of species but the persistence of unique behavioral traditions. This intersection of culture and conservation prompted researchers at the University of Arizona to develop an innovative and comprehensive initiative: The Animal Culture Database. Announced in a recent study published in the prestigious journal Scientific Data, this project collates decades of research on animal traditions into an openly accessible digital repository dedicated to cataloging vocal communications, mating displays, social play, and numerous other complex behaviors observed worldwide.

At its core, The Animal Culture Database employs a sophisticated, interactive global map that visualizes behavioral data linked to specific populations of animals. This spatially explicit tool enables scientists, educators, and conservationists to explore nuanced behaviors across species and regions—whether it’s dissecting the grooming rituals of gorillas, analyzing the breeding songs of avian species, or tracking the migratory patterns of whales. The database’s user-friendly interface offers multiple query options by species, behavior type, and geographic location, empowering exploratory research that bridges ethology, ecology, and conservation science.

The development of this ambitious database was spearheaded by Kiran Basava, a postdoctoral research associate at the University’s College of Information Science, in collaboration with a multidisciplinary team including faculty, postdoctoral scholars, and graduate students. Their collective effort involved an extensive meta-analysis of thousands of scientific publications, culminating in a rigorously curated selection of over a hundred studies that met stringent parameters for inclusion. Currently, the database encompasses detailed behavioral records from 30 mammalian species, 30 avian species, and one insect species, incorporating more than 120 studies in a structured, searchable format.

Such an initiative highlights the cross-disciplinary nature of contemporary animal behavior research. By integrating ecological data, social psychology, information science, and anthropology, the database not only synthesizes a fragmented body of literature but also facilitates novel insights into how culture evolves in the animal kingdom. One of the fundamental goals is to challenge the anthropocentric view that culture is a uniquely human trait, instead revealing a continuum of cultural complexity that extends across diverse taxa.

Moreover, ongoing efforts to expand the database promise to integrate findings from an additional six hundred studies. There are also plans to develop a participatory platform where external researchers can contribute new data, fostering a collaborative environment that evolves in tandem with advancements in the field. This open-source approach democratizes access to high-quality behavioral information, potentially accelerating the pace of discovery and application in areas ranging from wildlife management to behavioral ecology.

The resonance of this research touches on profound questions about the interconnectedness of humans and the natural world. By dissecting how animals learn from each other, adopt unique habits, and navigate social challenges, we gain a clearer understanding of the evolutionary roots of culture itself. Such knowledge informs ethical considerations about human impacts on ecosystems, emphasizing our obligation not only to preserve species but also to safeguard the intangible cultural heritage embedded in animal societies.

Furthermore, these insights carry practical implications for conservation strategies. As humans alter landscapes, introduce invasive species, or modify climate patterns, the traditional knowledge embedded in animal cultures may be disrupted or lost. Recognizing cultural diversity within species can refine translocation programs, helping conservationists ensure that moved animals can adapt and thrive socially in their new environments. It also strengthens arguments for protecting habitats that support not just biological but behavioral diversity.

The Animal Culture Database thus stands as a pioneering tool that merges data science with natural history. It catalyzes a new era of research where the behavioral richness of animals is mapped systematically and globally, offering a dynamic window into the complexity of life on Earth. As technology advances and datasets become richer, our capacity to decode the cultural tapestries woven through animal societies will deepen, inviting us to rethink long-held assumptions about intelligence, learning, and tradition beyond humanity.

In summation, this breakthrough initiative redefines how we perceive animal communication and culture, highlighting the diversity and adaptability of social behaviors across species. By recognizing and documenting the existence of nonhuman cultures, the scientific community not only challenges anthropocentrism but also enhances our ability to protect and understand the natural world in an era of rapid environmental change. The convergence of ethology, information science, and conservation biology represented by The Animal Culture Database marks a transformative step toward a more holistic appreciation and stewardship of animal life.

Subject of Research: Animals

Article Title: Mapping nonhuman cultures with the Animal Culture Database

News Publication Date: 17-Jun-2025

Web References: 10.1038/s41597-025-05315-y

References: The study published in Scientific Data

Keywords: Animal communication, Animal migration, Behavioral ecology, Mating behavior, Animal psychology, Cultural anthropology, Bird migration, Migration tracking, Data sets, Information processing

Credit: Institute for Basic Science

Credit: Michael McCarthy/UW Medicine