Why statistical learning researchers should study nonhuman animals.

Prediction in language comprehension across the lifespan

Large Language Models as Distributional Baselines for Language Tasks

The central contributions of this paper articulate the conditions under which distributional predictability threatens the internal validity of an experiment and provide concrete recommendations for how to control for this potential confound.

From Observation to Intervention: Memory in Brains and Large Language Models

Brains and large language models (LLMs) are fundamentally different memory systems, but they can be compared through shared functional questions: where memory-related information is represented, how partial cues recover broader associations, how new information is written or updated, and how memory-related states can be perturbed. In biological systems, these questions span synapses, neuronal ensembles, hippocampal-cortical interactions, and plasticity; in LLMs, they span weights, activations, context windows, retrieval systems, and external stores. The comparison is therefore functional and experimental rather than anatomical. Human studies reveal sparse concept responses, temporal binding, rapid association formation, episode-specific coding, and recall-related reactivation, but selective intervention remains limited. Rodent studies provide more selective causal access to learning-related ensembles, whereas human and macaque interventions usually affect broader circuits. LLMs lack lived episodic memory, yet they permit unusually direct and repeatable manipulation of internal states and stored information. We argue that this asymmetry creates a new opportunity. LLMs are not ahead in memory itself, but in experimental access. Their tools may help turn broad questions about retrieval, updating, persistence, reversibility, and unintended effects into sharper biological hypotheses. The productive bridge is to transfer experimental logic, not anatomical parts.

Shifts in learning dynamics drive developmental improvements in the acquisition of structured knowledge

It's cognition, with style: rethinking the biology of intelligence.

Intelligence has been attributed to a growing number of non-primate species, including birds, cephalopods, and jumping spiders. This review outlines key findings in animal intelligence, finding that there is significant support for the presence of a wide variety of intelligent capacities in these groups. Yet, how intelligence can be achieved with relatively small, non-mammalian brains remains an open question, and neuroscientific efforts have not established the relationship between brains and intelligence. Core to these research programs is the assumption that intelligence sits higher on a scale of complexity than other kinds of cognition, requiring greater neural resources. This review shows that, despite its ubiquity, there is little evidence available for this assumption, and that it does not align well with the goals of contemporary research on animal intelligence. I argue that, like other biological functions, cross-species comparisons of cognition are not usefully framed as a question of as better or worse, simpler or more complex. Cognitive differences between species are the result of differing cognitive styles, rather than gradations on a universal scale of cognition. Intelligence is the human cognitive style, and intelligent animals are those whose cognitive style bears a family resemblance to ours. This approach embraces the anthropocentrism of the concept of intelligence, while rejecting the idea that intelligence is a superior capacity. With this alternative intelligence concept in mind, investigations into animal intelligence and its neural basis could gain traction by shifting away from the search for greater absolute neural capacity in animals, instead approaching these smaller-brained species as models that reveal what is not required for intelligence.

Structured experience shapes strategy learning and neural dynamics in the medial entorhinal cortex.

Animals solve new, complex tasks by reusing and adapting prior knowledge. This flexibility depends not only on the content of experience but also on its structure. Early training curricula are especially important: poorly structured experiences can hinder abstraction and limit generalization. However, the neural mechanisms through which experience shapes future learning remain unclear. Here, we trained recurrent neural networks (RNNs) on an odor timing task used to study complex timing behavior in mice and then tested the model predictions with mouse behavior and medial entorhinal cortex recordings. Without structured early experience, both RNNs and mice developed rigid, error-prone strategies, whereas structured training promoted neural activity reflecting the task's temporal structure. Using dynamical systems analysis, we examined how different training curricula shaped network dynamics and whether these dynamics supported abstraction and generalization as task complexity increased. These findings demonstrate that the structure of prior experience governs how flexible, generalizable knowledge emerges in biological systems and computational models.