CoChat
Sign Up
Cited by 8Zebrafish Biomedical Research ApplicationsNeural dynamics and brain functionUndergraduate Neuroscience Education and Research

Artificial embodied circuits uncover neural architectures of vertebrate visuomotor behaviors

Open access
Liu, Xiangxiao, Loring, Matthew D., Zunino, Luca, Fouke, Kaitlyn E., Longchamp, François A., Bernardino, Alexandre, Ijspeert, Auke J., Naumann, Eva A.·2025·Science Robotics
DOI 10.1126/scirobotics.adv4408PMC PMC13077711
Open this paper in CoChat — free account

Abstract

To investigate the neural and biomechanical mechanisms of sensorimotor transformations, we constructed realistic neuromechanical simulations (simZFish) of the zebrafish optomotor response, a visual stabilization behavior. By computationally reproducing the body mechanics, physical body-water interactions, hydrodynamics, visual environments, and experimentally derived neural network architectures, we closely replicated the behavior of real larval zebrafish. Changing lens properties and retinal connectivity revealed why the lower posterior visual field drives optimal optomotor responses in the simZFish. In virtual rivers, the simZFish performed rheotaxis autonomously by using current-induced optic flow patterns as navigational cues. Last, experiments with a physical robot (ZBot) validated the role of embodied sensorimotor circuits in maintaining position in a real river.

Integrity signals

Registry

Crossref · 10.1126/scirobotics.adv4408

Cites

138 of 138 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

41 of 64 cited sources checked · 11 supported · 0 flagged · accuracy not yet measured

  • Interpretationreasoning

    These visuomotor computations can compensate for water current-induced displacement, especially where mechanosensory information from the lateral line organ is not reliable.

    “In principle, this suggests that these visuomotor computations can compensate for water current-induced displacement, especially in situations where mechanosensory information from the lateral line organ is not reliable (20).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [20] source

      The paper’s Abstract says flow-induced displacement is readily detectable as visual whole-field motion, and explains that a fish in a homogeneous current cannot detect the flow without cues from an external frame of reference. It does not establish that the particular visuomotor computations in the citing sentence compensate for displacement, or test their use when lateral-line information is unreliable. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Findingcitation

    When initialized with different heading directions, simZFish eventually oriented its heading direction against the flow, performing rheotaxis, as observed in real fish maintaining position in rivers.

    “When initialized with different heading directions, simZFish eventually oriented its heading direction against the flow, performing rheotaxis, as observed in real fish maintaining position in rivers (20).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [20] source

      The Abstract reports that real larval zebrafish oriented and swam against water flow in a laboratory tube, and that their ability to hold position improved with stronger flow gradients. It does not report observing fish maintain position in rivers; its discussion of small rivers concerns where the proposed mechanism might apply. The simZFish result is the citing authors’ own result, not a finding attributed to this source. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Findingcitation

    Systematic manipulations of sensory features, including simZFish's optical lens and retinal connectivity, revealed key factors driving neural architecture and response characteristics.

    “Further, we used systematic manipulations of sensory features, including simZFish’s optical lens and retinal connectivity, to reveal key factors driving neural architecture and response characteristics (31, 37).”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [31] source

      The Introduction (p. 4) reports mapping retinal inputs and reconstructing optic-flow-responsive neurons, supporting the retinal-connectivity and neural-architecture aspects. The paper does not describe simZFish, an optical-lens manipulation, or systematic manipulations of those simulated features. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [37] source

      The paper’s Summary and conclusions discusses multisensory rheotaxis and proposes future mapping of neural circuitry; its methodological-pitfalls section discusses sensory-deprivation treatments. Searches for simZFish, an optical lens, retinal connectivity, and systematic manipulations revealing neural architecture found no passage addressing the attributed result. The indexed passages provide no page numbers. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
  • Premisecitation

    These neurons exist in real zebrafish in at least eight mirror symmetric sets of direction-selective functional response types.

    “These neurons exist in real zebrafish in at least eight mirror symmetric sets of direction-selective functional response types (25).”
    Quote in the paperSubstantial claim1 of 1 sources checked
  • Findingown data

    simZFish spontaneous bout frequency is 0.6 Hz, compared to zebrafish 0.67 Hz ± 0.09 Hz.

    “To modulate the spontaneous bout frequency (simZFish: 0.6 Hz; zebrafish: 0.67 Hz +/− 0.09 Hz), the lPTs connect to symmetric neural nodes, representing the diencephalic nuclei of the medial longitudinal fasciculus (nMLF) known to control bout frequency (27), tail posture, and bout initialisation (41).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [27] source

      The Results section on whole-field visual motion reports that interbout duration changes with grating speed, but gives no spontaneous bout frequency of 0.67 ± 0.09 Hz. Searches for the stated frequency, its uncertainty, and spontaneous bout frequency found no passage establishing that value in our copy; the simZFish value is likewise not reported. (LLM, conf=0.75)

      Experimental — accuracy not yet measured
    • cites [41] source
  • May combine two claims. 51
  • Classified as probably not a substantial claim. 13

  • May combine two claims.

    Findingcitation

    simZFish1.0 successfully replicates key behavioral aspects of zebrafish OMR, partly due to its tuning to match behaviors for limited orthogonal stimuli.

    “The first iteration, simZFish1.0, successfully replicates key behavioral aspects of zebrafish OMR, partly due to its tuning to match behaviors for limited orthogonal stimuli (25).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [25] source

      The Results section, “Identifying Critical Dimensions of Functional Connectivity” (page unavailable), says the circuit-model architecture could accurately predict behavioral data, and the Methods describe fitting predictions to responses under ten stimulus conditions. Our copy does not identify that model as simZFish1.0 or establish that its success was partly due to tuning on limited orthogonal stimuli. (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Interpretationcitation

    Alternative neural circuits favoring lower anterior DSGC connectivity could be equally effective but would require downstream visuomotor circuits different from those experimentally observed.

    “Although it is plausible that alternative neural circuits favoring lower anterior DSGCs connectivity could be equally effective, they would require downstream visuomotor circuits different from those experimentally observed (25).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [25] source

      The Results section, “Identifying Critical Dimensions of Functional Connectivity” (page unavailable), says that multiple sets of connections can yield similar behavioral predictions, supporting the possibility of equally effective alternatives. It does not identify alternatives favoring lower anterior DSGC connectivity or say that they would require downstream visuomotor circuits different from those experimentally observed. (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Interpretationcitation

    The findings align well with biological studies showing that the OMR is most strongly evoked by motion in the lower posterior visual field, matching real pretectal receptive fields.

    “Notably, our findings align well with biological studies, which show that the OMR is most strongly evoked by motion in the lower posterior visual field (55), matching real pretectal receptive fields (31).”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [55] source

      The paper reports anisotropic color-processing circuits in the lower and outward visual field (Large-Scale Functional Anisotropies of the Inner Retina Match Natural Spectral Statistics, p. 10), but does not establish where motion most strongly evokes the OMR. Searches for optomotor responses and lower-posterior motion found no passage addressing that claim. The pretectal receptive-field statement has a separate citation and is not judged here. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
    • cites [31] source

      The citation to this paper is attached to the claim about pretectal receptive fields, not the separate claim about where motion most strongly evokes the OMR. The AF5 results (p. 9) show pretectal direction-selective responses to stimuli presented from the side and from below, but do not establish receptive fields preferentially covering the lower posterior visual field. The paper therefore supports pretectal motion sensitivity, not the asserted spatial match. (LLM, conf=0.77)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    lPTs collectively control movement direction by activating neurons of the anterior hindbrain, which suppresses behavioral oscillation via reciprocal inhibition.

    “As suggested by our previous modeling, lPTs collectively control the movement direction by activating neurons of the anterior hindbrain (aHB) (25, 30), which suppresses behavioral oscillation via reciprocal inhibition (Fig. 2E).”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [25] source
    • cites [30] source

      The ‘Motor related’ section maps neurons associated with locomotion, and the ‘Motor-off related’ section suggests that some dorsal-hindbrain neurons might inhibit motor output. Neither establishes that lPTs activate anterior-hindbrain neurons to control movement direction. A search for lPTs and reciprocal inhibition found no passage addressing that proposed circuit; page numbers are unavailable in our copy. (LLM, conf=0.74)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Current models of zebrafish visuomotor transformations rely on correlation-based neural activity, lack realistic retina or motor representations, omit neural connectivity manipulation, and exclude closed-loop observations of neural activation during behavior.

    “Current models of zebrafish visuomotor transformations rely on correlation-based neural activity, lack realistic representations of the retina or motor systems (25, 29), omit the ability to manipulate neural connectivity, and exclude closed-loop observations of neural activation during behavior, which limits the integrative investigation of the system within its sensory and physical environment.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [25] source

      The Discussion says the model ignores processing in the retina and spinal cord, and the Methods describe correlation-based neural activity maps. But the Results report both an experimental commissure ablation and its simulation in the model, so the claim that connectivity cannot be manipulated is not supported. The Introduction mentions closed-loop stimulus delivery, while the Methods describe neural imaging in restrained, paralyzed fish; these passages do not establish closed-loop neural observations during behavior. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
    • cites [29] source

      The Online methods (Algorithmic modeling; Anterior hindbrain network modeling, pages unavailable) describe simplified sensory inputs and swim-bout outputs, consistent with the claim that the model lacks realistic retina and motor-system representations. But the same methods say the authors varied connectivity weights and modeled visual feedback. The Two-photon calcium imaging methods also describe simultaneous neural imaging and behavioral readout with stimulus changes triggered by detected bouts, so the broader claims of omitted connectivity manipulation and excluded closed-loop observations are not supported. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Study design and conductcitation

    The simZFish platform generates a faithful virtual sensory environment with fluid dynamics, drag, and viscous forces, resulting in realistic locomotion behaviors.

    “The simZFish platform generates a faithful virtual sensory environment, complete with fluid dynamics, modeling drag, and viscous forces (Fig. S1C), resulting in realistic locomotion behaviors (Fig. S1D, E)(40).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [40] source

      The Methods section (page unavailable) describes a zebrafish swimming simulation that models fluid flow and viscous forces. The Results section (page unavailable) reports that its simulated body shapes matched recorded fish and could imitate self-propelled swimming. The paper does not identify this model as simZFish or describe a virtual sensory environment, so it does not support those parts of the claim. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Real larval zebrafish experience intermediate Reynolds numbers of 10-1000 and are influenced by both viscous and pressure drag.

    “ZBot operates in a turbulent regime, with a higher Reynolds number (1.35*105, at 0.15 m/s), experiencing lower viscous drag than the real larval zebrafish, which experience intermediate Reynolds numbers: 10–1000 (58) and are influenced by both viscous and pressure drag (59).”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [58] source

      The Introduction (page unavailable) mentions fish only in a discussion of optokinetic eye movements. Searches for larval zebrafish Reynolds numbers and for viscous and pressure drag found no passage in our copy addressing the hydrodynamic claim. In the citing sentence, the Reynolds-number range is attached to citation 58, while the drag claim is attached to this source as citation 59. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
    • cites [59] source

  • May combine two claims.

    Findingcitation

    Investigating ZBot's visual input shows the complexity of natural visual input, including turbidity, glare, and Snell's window, which limits the underwater field of view due to surface refraction.

    “Investigating the visual input of the ZBot demonstrates the complexity of natural visual input, including turbidity, glare, and Snell’s window, which limits the field of view underwater due to surface refraction (80).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [80] source

      The paper’s discussion of Figure 1 says that refraction at an air–water interface compresses the apparent horizon into a 97.2° Snell window for a fish in an experimental arena. It does not investigate ZBot’s natural visual input or address turbidity and glare, so those parts of the attributed claim are unsupported. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingcitation

    ZBot operates in a turbulent regime with a Reynolds number of 1.35×10^5 at 0.15 m/s and experiences lower viscous drag than real larval zebrafish.

    “ZBot operates in a turbulent regime, with a higher Reynolds number (1.35*105, at 0.15 m/s), experiencing lower viscous drag than the real larval zebrafish, which experience intermediate Reynolds numbers: 10–1000 (58) and are influenced by both viscous and pressure drag (59).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [58] source

      The Introduction discusses the optokinetic reflex, and the Results and Methods passages concern retinal activity and eye movements in mice; no page numbers are provided. Searches for ZBot, larval zebrafish, Reynolds number, and viscous or pressure drag found no passage addressing the attributed fluid-dynamics claim. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
    • cites [59] source

  • May combine two claims.

    Premisecitation

    simZFish and ZBot lack motor adaptation, such as adjusting speed to compensate for faster river flow, and are limited by their maximal swimming speed and bout frequency.

    “As simZFish and ZBot currently lack any ability for motor adaptation (60), such as adjusting their speed to compensate for faster river flow, they are limited by their maximal swimming speed and bout frequency.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [60] source

      The Abstract and Burst and coast behavior section examine swimming and coasting in living zebrafish larvae, while Two ways to maximize coast distance discusses physical limits on fish speed. Our copy does not address simZFish or ZBot, whether they can adapt their swimming to river flow, or limits on their bout frequency (searched: model names, motor adaptation, river flow, and bout frequency). (LLM, conf=0.78)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    Systematically altering connectivity between DSGCs and ePTs revealed that the lower posterior visual field is most effective, as only this connectivity strongly activated the downstream circuit to drive increased bout frequency.

    “Systematically altering the connectivity between DSGCs and ePTs revealed that the lower posterior visual field is most effective, as only this connectivity strongly activated the downstream circuit to drive increased bout frequency (Fig. S6E, F).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    simZFish and zebrafish respond to converging motion to both eyes with increased bout frequency, while diverging motion decreases bout frequency below spontaneous levels.

    “Comparing average histograms of bout angle distributions (Fig. 3E), simZFish and zebrafish respond to converging motion to both eyes, with increased bout frequency, while diverging motion decreases bout frequency below spontaneous levels.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    As predicted from optical principles, focal length affected the perceived visual projection onto the simulated retina, with a wider field of view corresponding to shorter focal lengths.

    “As predicted from optical principles, the focal length affected the perceived visual projection onto the simulated retina, with a wider field of view corresponding to shorter focal lengths.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Presenting these new stimuli to real zebrafish revealed substantial behavioral differences, suggesting mismatches between the real and simulated neural architectures.

    “Presenting these new stimuli to real zebrafish (Fig. 5C, middle), we observed substantial behavioral differences (Fig. S7A, B), suggesting mismatches between the real and simulated neural architectures.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Narrower angle lenses increase the optic flow magnitude of gratings moving orthogonally to the image plane of the camera or eye but have minimal effect on the optic flow magnitude of parallel stimuli.

    “Narrower angle lenses increase the optic flow magnitude of gratings moving orthogonally to the image plane of the camera or eye but have minimal effect on the optic flow magnitude of parallel stimuli (Fig. 4A).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Guided by simZFish predictions, neurons with previously undiscovered response profiles were identified through calcium imaging in live zebrafish brains and incorporated, leading to improved performance.

    “Guided by simZFish predictions, we also identified neurons with previously undiscovered response profiles through calcium imaging in live zebrafish brains and incorporated these newly discovered neural subtypes, leading to improved performance.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Some backward-selective neurons were suppressed by forward motion to the opposite eye, indicating a neural mechanism that reduces turning.

    “Mirroring the behavior, we discovered some backward-selective neurons were suppressed by forward motion to the opposite eye, indicating a neural mechanism that reduces turning (Fig. S8A).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    During early network tuning, connecting all DSGCs led to poor OMR performance, with no increased bout frequency in response to forward motion.

    “During early network tuning, we found that connecting all DSGCs led to poor OMR performance, with no increased bout frequency in response to forward motion.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In real zebrafish, monocular backward motion induced turning in the direction of the stimulated eye, indicating that the brain interprets this as a turning stimulus despite the absence of formal directional information.

    “However, in real zebrafish, monocular backward (xB, Bx) induced turning in the direction of the stimulated eye, indicating that the brain interprets this as a turning stimulus despite the absence of formal directional information.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In different flow velocities, simZFish maintains its position for speeds up to 0.3 cm/s, overcompensating for slower flow velocities.

    “In different flow velocities, simZFish maintains its position for speeds up to 0.3 cm/s, overcompensating for slower flow velocities (Fig. 3H).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Real zebrafish and simZFish1.0 strongly increased bout frequency to binocular forward motion, and monocular forward motion triggered about 50% of the bout rate.

    “As expected, real zebrafish and simZFish1.0 strongly increased bout frequency to binocular forward motion (FF, Fig. S7C–E), and monocular forward motion (xF, Fx) triggered ~50% of the bout rate.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    These behavioral results reveal complex interactions across eyes and motion directions in real zebrafish, which were not identified before and were absent from simZFish1.0.

    “These behavioral results reveal complex interactions across eyes and motion directions in real zebrafish, which were not identified before and, therefore, absent from simZFish1.0 (Fig. S7G).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In simZFish1.0, these stimuli strongly increased turning to the opposite side, predicting discrepancies in their neural architecture.

    “In simZFish1.0, these stimuli strongly increased turning to the opposite side, predicting discrepancies in their neural architecture (Fig. S7C).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Binocular leftward motion increased leftward bout frequency (0.027 Hz at −46°), increased biased forward bouts (0.089 Hz at −4°), and suppressed turns in the opposite direction (0.012 Hz at 34°).

    “Binocular leftward motion increased leftward bout frequency (0.027 Hz, at −46°), increased biased forward bouts (0.089 Hz, at −4°), and suppressed turns in the opposite direction (0.012 Hz, at 34°).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The physical ZBot validated these neural algorithms in the wild and helped stabilize its position in challenging natural environments with fast-flowing, turbid water.

    “The physical ZBot further validated these neural algorithms in the wild, contributing to stabilizing its position in challenging, natural environments with fast-flowing, turbid water conditions (Fig. 6), reinforcing its potential for studying sensorimotor systems in real-world environments.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In a stationary lake, ZBot showed swimming kinematics similar to real larval zebrafish, with realistic bout and glide behavior, a preset bout frequency of 0.09 Hz, and tail-beating frequencies of 1.5 Hz.

    “Despite the size differences, ZBot exhibited swimming kinematics like real larval zebrafish, demonstrating realistic bout and glide behavior with preset bout (0.09 Hz), and tail-beating frequencies (1.5 Hz).”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Real fish in natural rivers receive visual inputs that are richer than those that can be simulated, including differences in texture, turbidity, and lighting.
    • Rests on an assumption: The larger-scale ZBot design reasonably balances approximating larval zebrafish morphology with the technical requirements of a moving fish-like robot.

  • May combine two claims.

    Findingown data

    Neurons most critical for turning split into forward and backward selective types, with additional neurons responding to both and showing suppression to the contralateral eye.

    “Notably, neurons most critical for turning (oB, B, Mm, S), split into forward and backward selective types, with additional neurons responding to both and showing suppression to the contralateral eye (Supplementary Movie 7).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    simZFish accurately reproduces body-water interactions, neural circuits, and visual environments.

    “By synthesizing empirical neurobiological data into a physics-based simulator, we created simZFish, which accurately reproduces body-water interactions, neural circuits, and visual environments (Fig. 2).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    An expanded set of translational stimuli uncovered discrepancies between simZFish and real behavior and directly predicted functional neural subtypes with specific stimulus preferences and circuit layouts.

    “Crucially, these insights prompted us to use an expanded set of translational stimuli, uncovering discrepancies between the simZFish and real behavior, which directly predicted functional neural subtypes with specific stimulus preferences and circuit layouts (Fig. 5D).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    simZFish provides full control over sensory input and artificial neurons during free locomotion, allowing manipulation of neuronal properties and physical attributes such as lens characteristics or body proportions.

    “simZFish provides full control over sensory input and artificial neurons during free locomotion, offering virtually unlimited opportunities to manipulate neuronal properties and physical attributes like lens characteristics or body proportions (Supplementary Discussion).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Natural visual input was sufficient to orient ZBot upstream, with directed turn bouts adjusting heading and increased forward bout frequency when aligned with optic flow.

    “The remotely recorded neural activity demonstrates that the natural visual input is sufficient to orient ZBot upstream, adjusting heading direction with directed turn bouts and increasing forward bout frequency when aligned with the direction of optic flow (Fig. 6F).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In ZBot camera recordings, even in clear water, visual features above a certain elevation tend to be blurred and saturated by light, which emphasizes the importance of lower visual fields.

    “Recordings of the ZBot’s cameras, even in clear waters, emphasized the importance of lower visual fields, as visual features above a certain elevation tend to be blurred and saturated by light (Fig. 6C, S10A).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    These forward and backward-preferring subtypes were prevalent across all zebrafish, highlighting their importance in motion processing.

    “These forward and backward-preferring subtypes were prevalent across all zebrafish, highlighting their importance in motion processing (Fig. S8C–E).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Each ZBot tail-beating bout phase lasted about 5 seconds and completed five tail-beating cycles.

    “Each tail-beating bout phase lasted ~5 s, completing five tail-beating cycles (Fig. S9G–H).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S1 (asserted), S4 (based).

    Interpretationreasoning

    Lens morphology directly affects neural activation and sensory-driven behavior in identical neural networks.

    “Thus, lens morphology directly affects neural activation and sensory-driven behavior in identical neural networks (Fig. S6C, D).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    simZFish's circuits contribute to maintaining position in moving waters with visual input alone, effectively performing rheotaxis autonomously.

    “These simulations show that simZFish’s circuits contribute to maintaining position in moving waters with visual input alone, effectively performing rheotaxis autonomously.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    If evolutionary pressure favors minimizing neural connections, this configuration provides an effective solution, maximizing optic flow information for downstream neural circuits while minimizing wiring and computational load.

    “If evolutionary pressure favors minimizing neural connections, this configuration provides an effective solution, maximizing optic flow information for downstream neural circuits while minimizing wiring and computational load.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Evolutionary pressure favors minimizing the number of neural connections.

  • May combine two claims.

    Interpretationcitation

    Neural connectivity and ethologically relevant sensory input not only affect neural activity and resulting behavior but might even dictate neural architecture itself.

    “Practically, these results demonstrate that neural connectivity and ethologically relevant sensory input not only affect neural activity and resulting behavior but might even dictate neural architecture itself (57).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [57] source

  • May combine two claims.

    Premisecitation

    In real zebrafish, all DSGCs project contralaterally, activating neurons in the early pretectum.

    “As in real zebrafish (32), all DSGCs project contralaterally, activating neurons in the early pretectum (ePTs).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [32] source

  • May combine two claims.

    Interpretationown data

    The results underscore how embodiment shapes neural circuits and how mechanisms identified in idealized laboratory conditions can be implemented to lead to adaptive behaviors in the real world.

    “Together, our results underscore how embodiment shapes neural circuits and how mechanisms identified in idealized laboratory conditions can be implemented to lead to adaptive behaviors in the real world.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    These results validate the circuit functionality and provide evidence for OMR's purpose as a sensory-guided stabilization mechanism.

    “Importantly, these results not only validate the circuit functionality but also provide evidence for OMR’s purpose as a sensory-guided stabilization mechanism.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    simZFish generalized to unseen visual environments and performed rheotaxis robustly and autonomously.

    “Thus, this relatively complex behavior can emerge from sensing local optic flow, demonstrating that simZFish generalized to unseen visual environments, robustly and autonomously performing rheotaxis (36).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [36] source

  • May combine two claims.

    Premisecitation

    Simulations are cost-effective and allow systematic testing of modifications to the body or neural connectivity.

    “In contrast, simulations are cost-effective and allow for systematic testing of body or neural connectivity modifications (6, 61).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [6] source
    • cites [61]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data

    Incorporating these neurons provided a valuable strategy for reverse-engineering visuomotor systems and uncovered previously overlooked, behaviorally relevant neural computations.

    “Incorporating these neurons into simZFish improved behavioral accuracy and provided a valuable strategy for reverse-engineering visuomotor systems, uncovering previously overlooked, behaviorally relevant neural computations.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For all trials, ZBot was released with the same initial upstream heading direction of 0°.

    “For all trials, ZBot was released with the same initial upstream heading direction (0°).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    At least five repetitions of the 16 binocular and monocular orthogonal and parallel moving stimuli were presented.

    “At least five repetitions of the 16 binocular and monocular orthogonal and parallel moving stimuli were presented with a black bar underneath the fish.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    All river experiments were conducted within one hour on the same day to ensure comparable visibility and current velocity.

    “As the water flow in natural rivers varies rapidly, we conducted all experiments within one hour on the same day to ensure comparable visibility and current velocity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    This iterative process enhanced the simulated neural circuit's accuracy and realism, offering important insights into circuit function.

    “This iterative process enhanced the simulated neural circuit’s accuracy and realism, offering important insights into circuit function.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    simZFish served as a predictive framework that suggested new behavioral and neural experiments and led to identification of neural characteristics critical for biorealistic OMR behaviors.

    “Ultimately, simZFish served as a predictive framework that suggested novel behavioral and neural experiments, leading to the identification of neural characteristics critical for biorealistic OMR behaviors (Fig. 5).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    All live zebrafish experiments were approved by Duke University School of Medicine's animal care and use program (IACUC, Protocol Registry #A058-24-03).

    “All live zebrafish experiments were approved by Duke University School of Medicine’s animal care and use program (IACUC, Protocol Registry #A058-24-03).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Current stabilization methods in floating robots such as quadcopters and underwater vehicles often use multiple cameras pointing downward.

    “The current methods for stabilization in floating robots, such as quadcopters and underwater vehicles often use multiple cameras pointing downward (81).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [81] source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Of 29 trials, 14 were included in the final analysis; 16 were excluded because of collisions with the river edge (4), swimming upstream out of the drone camera view (3), and Wi-Fi communication failure (6).

    “Of 29 trials, we included 14 trials in the final analysis because the other 16 trials were corrupted due to collisions with the river edge (4 trials), swimming upstream out of the drone camera view (3 trials), and failure of Wi-Fi communication (6 trials).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    ZBot is unique in using a biorealistic, neurobiologically derived artificial neural architecture.

    “Nonetheless, ZBot is unique in using a biorealistic, neurobiologically derived artificial neural architecture.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    River experiments were conducted only if it had not rained in the past five days, to avoid muddy water decreasing visibility.

    “As rain muddied the water, decreasing visibility, experiments were only conducted if it had not rained in the past five days.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The zebrafish-inspired artificial neural architecture with just two laterally positioned eyes offers a bio-inspired strategy for stabilizing fish-like robots in fluid flows and reduces cost and weight.

    “Our zebrafish-inspired artificial neural architecture, featuring just two laterally positioned eyes, offers a bio-inspired strategy for stabilizing fish-like robots in fluid flows, thereby reducing cost and weight.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Zebrafish were removed from further analysis if there were imaging artifacts or if the pretectal population response did not show consistent activation by whole-field visual stimuli.

    “We removed zebrafish from further analysis if there were imaging artifacts or if the population response in the pretectum did not show consistent activation by whole-field visual stimuli.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Zebrafish were removed from further analysis if there were tracking artifacts or if the fish did not perform at least one trial of each visual stimulus repetition.

    “We removed zebrafish from further analysis if there were tracking artifacts or the fish did not perform at least one trial of each visual stimulus repetition.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    For all zebrafish behavior, the study met or exceeded the field's standard for number of experimental animals (>16) and trial repetitions (>10, >25 s trial length); for imaging studies (>7 animals), stimulus repetitions (>5).

    “For all zebrafish behavior, we met or exceeded the field’s standard for the number of experimental animals (>16) and trial repetitions (>10, >25 s trial length); for imaging studies (> 7 animals), stimulus repetitions (>5).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    Because viscous drag exceeds pressure drag during gliding, larval zebrafish experience higher drag while gliding and therefore have proportionally shorter gliding distances than ZBot.

    “Since viscous drag is higher than pressure drag during the gliding phase, the comparatively small body of larval zebrafish experiences higher drag while gliding (59), resulting in proportionally shorter gliding distances compared to the ZBot (Fig. S9I).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [59] source
59 of 59

  • The authors state this as an assumption.

    To grasp how neural circuits work, one must test the whole brain-body system in settings that resemble the real world.

    “Understanding of their function requires integrative brain-body testing in realistic contexts.”

  • The authors state this as an assumption.

    To model real animals' neural dynamics accurately, one must validate sensorimotor loops in complex physical environments and account for nuanced behavioral outcomes.

    “However, accurate modeling of neural dynamics observed in real animals requires validation of sensorimotor loops within the complexities of the physical world and associated nuanced behavioral consequences.”

  • The authors state this as an assumption.

    Evolutionary pressure favors minimizing the number of neural connections.

    “If evolutionary pressure favors minimizing neural connections, this configuration provides an effective solution, maximizing optic flow information for downstream neural circuits while minimizing wiring and computational load.”

  • The paper relies on this without support.

    Real fish in natural rivers receive visual inputs that are richer than those that can be simulated, including differences in texture, turbidity, and lighting.

    “For instance, real fish in natural rivers experience far richer visual inputs than we can simulate, including variations in texture, turbidity, and lighting.”

  • The paper relies on this without support.

    Real-world fluid dynamics pose challenges such as turbulence and irregular flow patterns.

    “Real-world fluid dynamics present challenges, such as turbulences and irregular flow patterns.”

  • The paper relies on this without support.

    Current technology does not allow construction of robots as small as the larval zebrafish body while including the necessary hardware.

    “Although the ZBot would ideally replicate the small size (~4 mm) of the larval zebrafish body, current technology does not permit the construction of such small robots with the required hardware (cameras, controller, motors, electronic boards, batteries, waterproofing, etc.).”

  • The paper relies on this without support.

    The larger-scale ZBot design reasonably balances approximating larval zebrafish morphology with the technical requirements of a moving fish-like robot.

    “Thus, our larger-scale ZBot (~ 80 cm) balances key features of real larval zebrafish with technological requirements (Fig. S9A).”
    “ZBot’s length (~80 cm) and weight (2.7 kg) represent reasonable trade-offs between approximating real larval zebrafish morphological features and satisfying technical requirements to equip a moving fish-like robot with cameras, electronic boards, batteries, motors, and waterproofing.”
7 of 7

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

  • Corruption subcategories do not add to the stated total

    Major

    The listed reasons for excluding trials are 4 collisions, 3 out of view, and 6 Wi-Fi failures. These add to 13, not the stated 16 corrupted trials.

    Reported 16 · recomputed 13

  • Included and excluded ZBot trials do not sum to total trials

    Major

    The methods state that 29 trials were run, 14 were included, and the other 16 were corrupted. But 14 plus 16 equals 30, not 29, so the trial accounting is internally inconsistent.

    Reported 29 · recomputed 30

2 of 2

Cites

Flagged first, then by citations

Research works printed but not linked: 3

  1. 10.1038/s41586-024-07633-4 · DOI · Paper page
  2. J. Merel D. Aldarondo J. Marshall Y. Tassa G. Wayne B. Olveczky “Deep neuroethology of a virtual rodent” in ICLR 2020: The Eighth International Conference on Learning Representations (ICLR 2019) pp. 1–20; https://openreview.net/forum?id=SyxrxR4KPS.
  3. D. E. Rumelhart R. Durbin R. Golden Y. Chauvin “Backpropagation: The basic theory” in Backpropagation: Theory Architectures and Applications (Lawrence Erlbaum Associates Inc 1995) pp. 1–34.

Cited by

Sorted by citations

Related

Authors
Liu, X., Loring, M. D., Zunino, L., Fouke, K. E., Longchamp, F. A., Bernardino, A., Ijspeert, A. J., & Naumann, E. A.
Affiliations
  • École Polytechnique Fédérale de Lausanne (CH)
  • Duke University (US)
  • Instituto Superior Técnico (PT)
Corresponding author
Naumann, Eva A.
ORCID
Venue
Science Robotics
Publisher
American Association for the Advancement of Science (AAAS)
Published
2025-10-15
Volume
10
Issue
107
Pages
eadv4408-eadv4408
ISSN
2470-9476 (ISSN-L)
DOI
10.1126/scirobotics.adv4408
PMC
PMC13077711
OpenAlex
W4415221481
Cited by
8
References
130
Citations by year
2026: 8
FWCI
3.97
Type
journal article
Language
en
Open access
Locations
  • Science Robotics (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
Funders
  • National Eye Institute (R01EY033845)
  • National Institute of Neurological Disorders and Stroke (R34NS116738, RF1NS128895, R01NS128895)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cell Biology › Zebrafish Biomedical Research Applications
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Undergraduate Neuroscience Education and Research
Keywords
sensorimotor circuits, embodiment, zebrafish larvae, rheotaxis, optical flow, retinal connectivity, receptive fields, fluid dynamics, neural circuit architecture
MeSH terms
Animals, Behavior, Animal, Biomechanical Phenomena, Brain, Computer Simulation, Larva, Nerve Net, Robotics, Swimming, Zebrafish, Neural Networks, Computer, Rivers, Hydrodynamics, Optic Flow

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read under definition version 3 (extract-v3) on 2026-09-28.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review