共找到 20 条结果
Amphibious legged robots inspired by salamanders are promising in applications in complex amphibious environments. However, despite the significant success of training controllers that achieve diverse locomotion behaviors in conventional quadrupedal robots, most salamander robots relied on central-pattern-generator (CPG)-based and model-based coordination strategies for locomotion control. Learning unified joint-level whole-body control that reliably transfers from simulation to highly articulated physical salamander robots remains relatively underexplored. In addition, few legged robots have tried learning-based controllers in amphibious environments. In this work, we employ Reinforcement Learning to map proprioceptive observations and commanded velocities to joint-level actions, allowing coordinated locomotor behaviors to emerge. To deploy these policies on hardware, we adopt a system-level real-to-sim matching and sim-to-real transfer strategy. The learned controller achieves stable and coordinated walking on both flat and uneven terrains in the real world. Beyond terrestrial locomotion, the framework enables transitions between walking and swimming in simulation, highlighting a
Robots benefit from sensory information to coordinate body movement, gain robustness against perturbations, and transition between different modes to adapt to various terrains. However, few amphibious robots can sense interactions with both terrestrial and aquatic environments. In this paper, we present a solution that uses Hall-effect sensors to sense foot contact forces and lateral hydrodynamic forces on a salamander-inspired amphibious robot. With two bus lines, the robot can simultaneously acquire this exteroceptive information at more than 500 Hz and proprioceptive information, such as joint positions and loads, at 100 Hz. The Hall-effect sensors used are compact, making them suitable for embedding in multiple positions within a robot, and exhibit high sensitivity to small forces. Moreover, because the sensor can be positioned separately from the measured object, waterproofing can be implemented with relative ease. Our tests demonstrate the robot's capabilities in traversing amphibious environments and its potential in using feedback control for more complex locomotion tasks.
Chain-of-Thought (CoT) prompting enhances the math reasoning capability of large language models (LLMs) to a large margin. However, the mechanism underlying such improvements remains unexplored. In this paper, we present \textbf{SalaMAnder} (\textbf{S}h\textbf{a}p\textbf{l}ey-b\textbf{a}sed \textbf{M}athematical Expression \textbf{A}ttribution a\textbf{nd} M\textbf{e}t\textbf{r}ic), a theoretically grounded methodology as well as a mathematically rigorous evaluation metric for quantifying component-level contributions in few-shot CoT reasoning. Concretely, we leverage the Shapley value for mathematical expression attribution and develop an efficient stratified sampling algorithm that significantly reduces the computational complexity. Besides, we develop the \textbf{CoSP} (\textbf{C}ardinality \textbf{o}f \textbf{S}hapley \textbf{P}ositives) metric through covariance analysis. Comprehensive validation across popular LLM models and diverse mathematical benchmarks demonstrates that the CoSP metric within our SalaMAnder framework exhibits a robust monotonic correlation with model performance, not only providing theoretical explanations for the empirical success of existing few-shot Co
Studying the interaction between core-collapse supernova remnants (SNRs) and their surrounding environments is essential to understanding the mechanism for energy transfer to the interstellar medium (ISM) and the intrinsic physical properties of these remnants. In this paper, we focus on G309.8-2.6. Our new observations reveal that this object includes an SNR shell with a relic pulsar wind nebula (PWN) that extends well beyond the emission that has been previously observed in X-rays. We present new radio continuum and polarization images of G309.8-2.6 from the Evolutionary Map of the Universe (EMU) and Polarization Sky Survey of the Universe's Magnetism (POSSUM) surveys with the Australian Square Kilometre Array Pathfinder (ASKAP). The images reveal the complex and peculiar morphology of G309.8-2.6. The linear polarization displays an atypical S-shaped morphology and a highly ordered magnetic field. The rotation measure (RM) map shows a large-scale gradient or possible sign reversal, depending on the foreground RM. We reprocessed archival X-ray observations from Chandra and eROSITA, and retrieved archival H$α$ and infrared observations. We performed a joint analysis of the multiwav
Salamander-like robots are designed inspired by the skeletal structure of their biological counterparts. However, existing controllers cannot fully exploit these morphological features and largely rely on predefined patterns or joint trajectories, which prevents the generation of diverse and flexible gaits and limits their applicability in real-world scenarios. In this paper, we propose a phase-based learning framework that enables the robot to acquire a diverse repertoire of gaits without using reference motions. Each body part is controlled by a phase variable capable of forward and backward evolution, with a phase coverage reward to promote the exploration of the leg phase space. Additionally, morphological symmetry of the robot is incorporated via data augmentation, improving sample efficiency and enforcing both motion-level and task-level symmetry in learned behaviors. Extensive experiments show that the robot successfully acquires 22 representative gaits exhibiting both dynamic and symmetric movements, demonstrating the effectiveness of the proposed learning framework.
Very low levels of genetic diversity have been reported in vertebrates with large genomes, notably salamanders and lungfish [1-3]. Interpreting differences in heterozygosity, which reflects genetic diversity in a population, is complicated because levels of heterozygosity vary widely between conspecific populations, and correlate with many different physiological and demographic variables such as body size and effective population size. Here we return to the question of genetic variability in salamanders and report on the relationship between evolutionary rates and genome sizes in five different salamander families. We found that rates of evolution are exceptionally low in salamanders as a group. Evolutionary rates are as low as those reported for cartilaginous fish, which have the slowest rates recorded so far in vertebrates [4]. We also found that, independent of life history, salamanders with the smallest genomes (14 pg) are evolving at rates two to three times faster than salamanders with the largest genomes (>50 pg). After accounting for evolutionary duration, we conclude that more recently evolved species have correspondingly smaller genomes compared to older taxa and conc
An increasing number of studies have demonstrated correlations between climate trends and body size change of organisms. In many cases, climate might be expected to influence body size by altering thermoregulation, energetics or food availability. However, observed body size changes can result from ecological processes (e.g., growth, selection, population dynamics), yet may also be due to imperfect observation. We used two extensive datasets to evaluate alternative hypotheses for recently reported changes in the observed body size of plethodontid salamanders. We found that mean adult body size of salamanders can be highly sensitive to survey conditions, particularly rainfall, such that smaller individuals are more likely to be sampled under dry conditions. This systematic bias in the detection of individuals across a range in body size would result in a signature of body size reduction in relation to reported climate trends when it is simply observation error. We also identify considerable variability in body size distributions among years that shows correspondence with rainfall. This suggests that annual variation in growth or shifting population age structure could also result in
Salamanders (urodela) have among the largest vertebrate genomes, ranging in size from 10 to over 80 pg. The urodela are divided into ten extant families each with a characteristic range in genome size. Although changes in genome size often occur randomly and in the absence of selection pressure, non-random patterns of genome size variation are evident among specific vertebrate lineages. Here we report that genome size in salamander families varies inversely with species richness and other ecological factors: clades that began radiating earlier (older crown age) tend to have smaller genomes, higher levels of diversity and larger geographical ranges. These observations support the hypothesis that urodel families with larger genomes either have a lower propensity to diversify or are more vulnerable to extinction than families with smaller genomes.
It is well known that the classical diagram lemmas of homological algebra for abelian groups can be generalized to non-abelian group-like structures, such as groups, rings, algebras, loops, etc. In this paper we establish such a generalization of the "salamander lemma" due to G. M. Bergman, in a self-dual axiomatic context (developed originally by Z. Janelidze), which applies to all usual non-abelian group-like structures and also covers axiomatic contexts such as semi-abelian categories in the sense of G. Janelidze, L. Marki and W. Tholen and exact categories in the sense of M. Grandis.
Generative models are increasingly used to capture correlations in many-body systems, but the representations they learn remain largely opaque to physical interpretation. Here, we establish an intuitive criterion that quantifies the capacity of a variational autoencoder (VAE) to faithfully reconstruct the joint probability distribution of a many body system. In a nutshell, a bound on the VAE capacity is obtained by comparing the rate of the latent channel to the bipartite mutual information of the data. Using this bound, we show that the conditionally independent decoder of any successful VAE is structurally identical to a finite-size mean-field factorization. Hence, a successful reconstruction is direct evidence for a latent mean-field theory and the microscopic parameters of that theory can be read off the trained decoder. We validate these conclusions on a hierarchy of solvable models with scalar (Curie-Weiss), vector (Hopfield) and tensor (Maier-Saupe) order parameters, recovering the full Hopfield pattern matrix from equilibrium samples alone. We find that, when applied to Salamander retinal recordings, a two-latent VAE reproduces the population statistics with only two effect
Automated individual animal re-identification is essential for large-scale biodiversity monitoring; however, field imagery complicates separating identity cues from nuisance variation in pose, illumination, background, resolution, and species-specific morphology. The DS@GT ARC submission to AnimalCLEF 2026 introduces a multi-species image-clustering system for re-identifying Eurasian lynx, fire salamanders, loggerhead sea turtles, and Texas horned lizards. Instead of relying on a single descriptor or nearest-neighbor retrieval, this approach formulates re-identification as species-aware graph construction over candidate image pairs. The pipeline integrates tailored preprocessing, global candidate retrieval, LightGlue-based local verification with multiple keypoint families, LightGBM pair scoring, conservative edge admission, and Leiden community detection. This design directly addresses a primary failure mode of clustering-based re-identification: high-scoring false pairs that act as bridge edges and merge distinct individuals through transitive closure. Across species, ablation studies demonstrate that local feature support, foreground-aware preprocessing, and species-specific bac
Sprawling locomotion in vertebrates, particularly salamanders, demonstrates how body undulation and spinal mobility enhance stability, maneuverability, and adaptability across complex terrains. While prior work has separately explored biologically inspired gait design or deep reinforcement learning (DRL), these approaches face inherent limitations: open-loop gait designs often lack adaptability to unforeseen terrain variations, whereas end-to-end DRL methods are data-hungry and prone to unstable behaviors when transferring from simulation to real robots. We propose a hybrid control framework that integrates Hildebrand's biologically grounded gait design with DRL, enabling a salamander-inspired quadruped robot to exploit active spinal joints for robust crawling motion. Our evaluation across multiple robot configurations in target-directed navigation tasks reveals that this hybrid approach systematically improves robustness under environmental uncertainties such as surface irregularities. By bridging structured gait design with learning-based methodology, our work highlights the promise of interdisciplinary control strategies for developing efficient, resilient, and biologically info
This work explores Liquid Time-Constant Networks (LTCs) and Closed-form Continuous-time Networks (CfCs) for modeling retinal ganglion cell activity in tiger salamanders across three datasets. Compared to a convolutional baseline and an LSTM, both architectures achieved lower MAE, faster convergence, smaller model sizes, and favorable query times, though with slightly lower Pearson correlation. Their efficiency and adaptability make them well suited for scenarios with limited data and frequent retraining, such as edge deployments in vision prosthetics.
The capability of effectively moving on complex terrains such as sand and gravel can empower our robots to robustly operate in outdoor environments, and assist with critical tasks such as environment monitoring, search-and-rescue, and supply delivery. Inspired by the Mount Lyell salamander's ability to curl its body into a loop and effectively roll down {\Revision hill slopes}, in this study we develop a sand-rolling robot and investigate how its locomotion performance is governed by the shape of its body. We experimentally tested three different body shapes: Hexagon, Quadrilateral, and Triangle. We found that Hexagon and Triangle can achieve a faster rolling speed on sand, but exhibited more frequent failures of getting stuck. Analysis of the interaction between robot and sand revealed the failure mechanism: the deformation of the sand produced a local ``sand incline'' underneath robot contact segments, increasing the effective region of supporting polygon (ERSP) and preventing the robot from shifting its center of mass (CoM) outside the ERSP to produce sustainable rolling. Based on this mechanism, a highly-simplified model successfully captured the critical body pitch for each ro
We present a novel approach to neural response prediction that incorporates higher-order operations directly within convolutional neural networks (CNNs). Our model extends traditional 3D CNNs by embedding higher-order operations within the convolutional operator itself, enabling direct modeling of multiplicative interactions between neighboring pixels across space and time. Our model increases the representational power of CNNs without increasing their depth, therefore addressing the architectural disparity between deep artificial networks and the relatively shallow processing hierarchy of biological visual systems. We evaluate our approach on two distinct datasets: salamander retinal ganglion cell (RGC) responses to natural scenes, and a new dataset of mouse RGC responses to controlled geometric transformations. Our higher-order CNN (HoCNN) achieves superior performance while requiring only half the training data compared to standard architectures, demonstrating correlation coefficients up to 0.75 with neural responses (against 0.80$\pm$0.02 retinal reliability). When integrated into state-of-the-art architectures, our approach consistently improves performance across different sp
The characterization of network and biophysical properties from neural spiking activity is an important goal in neuroscience. A framework that provides unbiased inference on causal synaptic interaction and single neural properties has been missing. Here we applied the stochastic dynamics extension of Maximum Entropy -- the Maximum Caliber Principle -- to infer the transition rates of network states. Effective synaptic coupling strength and neuronal response functions for various network motifs can then be computed. The inferred minimal model also enables leading-order reconstruction of inter-spike interval distribution. Our method is tested with numerical simulated spiking networks and applied to data from salamander retina.
Decoding images from brain activity has been a challenge. Owing to the development of deep learning, there are available tools to solve this problem. The decoded image, which aims to map neural spike trains to low-level visual features and high-level semantic information space. Recently, there are a few studies of decoding from spike trains, however, these studies pay less attention to the foundations of neuroscience and there are few studies that merged receptive field into visual image reconstruction. In this paper, we propose a deep learning neural network architecture with biological properties to reconstruct visual image from spike trains. As far as we know, we implemented a method that integrated receptive field property matrix into loss function at the first time. Our model is an end-to-end decoder from neural spike trains to images. We not only merged Gabor filter into auto-encoder which used to generate images but also proposed a loss function with receptive field properties. We evaluated our decoder on two datasets which contain macaque primary visual cortex neural spikes and salamander retina ganglion cells (RGCs) spikes. Our results show that our method can effectively
Neurons subject to a common non-stationary input may exhibit a correlated firing behavior. Correlations in the statistics of neural spike trains also arise as the effect of interaction between neurons. Here we show that these two situations can be distinguished, with machine learning techniques, provided the data are rich enough. In order to do this, we study the problem of inferring a kinetic Ising model, stationary or nonstationary, from the available data. We apply the inference procedure to two data sets: one from salamander retinal ganglion cells and the other from a realistic computational cortical network model. We show that many aspects of the concerted activity of the salamander retinal neurons can be traced simply to the external input. A model of non-interacting neurons subject to a non-stationary external field outperforms a model with stationary input with couplings between neurons, even accounting for the differences in the number of model parameters. When couplings are added to the non-stationary model, for the retinal data, little is gained: the inferred couplings are generally not significant. Likewise, the distribution of the sizes of sets of neurons that spike si
The notion of L-homologies (of double complexes) as proposed in this paper extends the notion of classical horizontal and vertical homologies, along with two other new homologies introduced in the homological diagram lemma called salamander lemma. We enumerate all L-homologies associated with an object of a double complex and provide new examples of exact sequences. We describe a classification problem of these exact sequences. We study two poset structures on these L-homologies; one of them determines the trivialities of horizontal and vertical homologies of an object in terms of other L-homologies of that object, whereas the second structure shows the significance of the two homologies introduced in salamander lemma. Finally, we prove the existence of a faithful amnestic Grothendieck fibration from the category of L-homologies to a category consisting of objects and morphisms of a given double complex.
Computation in the brain involves multiple types of neurons, yet the organizing principles for how these neurons work together remain unclear. Information theory has offered explanations for how different types of neurons can optimize the encoding of different stimulus features. However, recent experiments indicate that separate neuronal types exist that encode the same stimulus features, but do so with different thresholds. Here we show that the emergence of these types of neurons can be quantitatively described by the theory of transitions between different phases of matter. The two key parameters that control the separation of neurons into subclasses are the mean and standard deviation of noise levels among neurons in the population. The mean noise level plays the role of temperature in the classic theory of phase transitions, whereas the standard deviation is equivalent to pressure, in the case of liquid-gas transitions, or to magnetic field for magnetic transitions. Our results account for properties of two recently discovered types of salamander OFF retinal ganglion cells, as well as the absence of multiple types of ON cells. We further show that, across visual stimulus contr