Empirically, S-TEC restructures meaningfully the within-and between-class representations.

A central question in both cognitive science and artificial intelligence is how humans and artificial systems can acquire competencies for language and motor command in a co-developmental manner, despite having access to only limited learning experiences. This question is exemplified in human infants, who achieve remarkable generalization with sparse input. This is a stark contrast to large-scale models which rely on massive training corpora, to reach similar capabilities. This raises the issue of what mechanisms enable such efficient developmental learning. From the perspective of developmental psychology, infants acquire language through rich interaction with their embodied environments. T omasello's "verb-island" hypothesis argues that children initially learn verbs in specific, isolated contexts before generalizing across broader linguistic structures (1). He also emphasized the importance of embodiment in language acquisition, suggesting that grounding linguistic symbols in sensorimotor experiences is fundamental to language learning (2).

Specifically, the brain also transforms the environment through efference, i.e.

Abstract-- We introduce PhysicalAgent, an agentic framework for robotic manipulation that integrates iterative reasoning, diffusion-based video generation, and closed-loop execution. Given a textual instruction, our method generates short video demonstrations of candidate trajectories, executes them on the robot, and iteratively re-plans in response to failures. This approach enables robust recovery from execution errors. We evaluate PhysicalAgent across multiple perceptual modalities (egocentric, third-person, and simulated) and robotic embodiments (bimanual UR3, Unitree G1 humanoid, simulated GR1), comparing against state-of-the-art task-specific baselines. Experiments demonstrate that our method consistently outperforms prior approaches, achieving up to 83% success on human-familiar tasks. Physical trials reveal that first-attempt success is limited (20-30%), yet iterative correction increases overall success to 80% across platforms. These results highlight the potential of video-based generative reasoning for general-purpose robotic manipulation and underscore the importance of iterative execution for recovering from initial failures. Our framework paves the way for scalable, adaptable, and robust robot control. The rapid progress of large foundation models has transformed the design of AI agents.

Laboratory for Intelligent Systems and Informatics, Department of Mechano-Informatics, Graduate School of Information Science and Technology, TheUniversity of Tokyo, Bunkyo-ku, Tokyo 113-8656, Japan (Dated: August 21, 2025) Neuronal spikes directly drive muscles and endow animals with agile movements, but applying the spike-based control signals to actuators in artificial sensor-motor systems inevitably causes a collapse of learning. We developed a system that can vary the number of independent synaptic bundles in sensor-to-motor connections. This paper demonstrates the following four findings: (i) Learning collapses once the number of motor neurons or the number of independent synaptic bundles exceeds a critical limit. The functions of spikes remain largely unknown. Identifying the parameter range in which learning systems using spikes can be constructed will make it possible to study the functions of spikes that were previously inaccessible due to the difficulty of learning.

This paper presents a learned model to predict the robot-centric velocity of an underwater robot through dynamics-aware proprioception. The method exploits a recurrent neural network using as inputs inertial cues, motor commands, and battery voltage readings alongside the hidden state of the previous time-step to output robust velocity estimates and their associated uncertainty. An ensemble of networks is utilized to enhance the velocity and uncertainty predictions. Fusing the network's outputs into an Extended Kalman Filter, alongside inertial predictions and barometer updates, the method enables long-term underwater odometry without further exteroception. Furthermore, when integrated into visual-inertial odometry, the method assists in enhanced estimation resilience when dealing with an order of magnitude fewer total features tracked (as few as 1) as compared to conventional visual-inertial systems. Tested onboard an underwater robot deployed both in a laboratory pool and the Trondheim Fjord, the method takes less than 5ms for inference either on the CPU or the GPU of an NVIDIA Orin AGX and demonstrates less than 4% relative position error in novel trajectories during complete visual blackout, and approximately 2% relative error when a maximum of 2 visual features from a monocular camera are available.

AI's significant recent advances using general-purpose circuit computations offer a potential window into how the neocortex and cerebellum of the brain are able to achieve a diverse range of functions across sensory, cognitive, and motor domains, despite their uniform circuit structures. However, comparing the brain and AI is challenging unless clear similarities exist, and past reviews have been limited to comparison of brain-inspired vision AI and the visual neocortex. Here, to enable comparisons across diverse functional domains, we subdivide circuit computation into three elements -- circuit structure, input/outputs, and the learning algorithm -- and evaluate the similarities for each element. With this novel approach, we identify wide-ranging similarities and convergent evolution in the brain and AI, providing new insights into key concepts in neuroscience. Furthermore, inspired by processing mechanisms of AI, we propose a new theory that integrates established neuroscience theories, particularly the theories of internal models and the mirror neuron system. Both the neocortex and cerebellum predict future world events from past information and learn from prediction errors, thereby acquiring models of the world. These models enable three core processes: (1) Prediction -- generating future information, (2) Understanding -- interpreting the external world via compressed and abstracted sensory information, and (3) Generation -- repurposing the future-information generation mechanism to produce other types of outputs. The universal application of these processes underlies the ability of the neocortex and cerebellum to accomplish diverse functions with uniform circuits. Our systematic approach, insights, and theory promise groundbreaking advances in understanding the brain.

This paper presents a method to recover quadrotor UAV from a throw, when no control parameters are known before the throw. We leverage the availability of high-frequency rotor speed feedback available in racing drone hardware and software to find control effectiveness values and fit a motor model using recursive least squares (RLS) estimation. Furthermore, we propose an excitation sequence that provides large actuation commands while guaranteeing to stay within gyroscope sensing limits. After 450ms of excitation, an INDI attitude controller uses the 52 fitted parameters to arrest rotational motion and recover an upright attitude. Finally, a NDI position controller drives the craft to a position setpoint. The proposed algorithm runs efficiently on microcontrollers found in common UAV flight controllers, and was shown to recover an agile quadrotor every time in 57 live experiments with as low as 3.5m throw height, demonstrating robustness against initial rotations and noise. We also demonstrate control of randomized quadrotors in simulated throws, where the parameter fitting RMS error is typically within 10% of the true value.

The higher stage is a neural network model that treats the cerebellum as an array of adjustable motor pattern generators. This network uses sensory input to preset and to trigger elemental pattern generators and to evaluate their performance. The actual patterned outputs, however, are produced by intrin(cid:173) sic circuitry that includes recurrent loops and is thus capable of self-sustained activity. These patterned outputs are sent as motor commands to local feedback systems called motor servos. Overall control is thus achieved in two stages: (1) an adaptive cerebellar network generates an array of feedforward motor commands and (2) a set of local feedback systems translates these commands into actual movements.

We propose a paradigm for modeling speech production based on neural networks. We focus on characteristics of the musculoskeletal system. Using real physiological data - articulator movements and EMG from muscle activity(cid:173) a neural network learns the forward dynamics relating motor commands to muscles and the ensuing articulator behavior. After learning, simulated perturbations, were used to asses properties of the acquired model, such as natural frequency, damping, and interarticulator couplings. Finally, a cascade neural network is used to generate continuous motor commands from a sequence of discrete articulatory targets.