Tianhe Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Tianhe Wang has a total of 51 co-authors. Here are the top co-authors by number of shared publications:
Richard B. Ivry
Department of Psychology and Neuroscience, University of California, Berkeley, California
18 shared publications- Jonathan Sanching Tsay
Department of Psychology, Carnegie Mellon University
10 shared publications Jordan A. Taylor
Princeton Neuroscience Institute, Princeton, New Jersey
9 shared publications- Richard B. Ivry
University of California, Berkeley
8 shared publications Tony Lam
University of California
8 shared publications- Guy Avraham
Department of Psychology and Helen Wills Neuroscience Institute, University of California
6 shared publications
Latest Publications
- Open Access
Development of appropriate sensorimotor recalibration for a de novo skill
Implicit adaptation operates automatically to ensure that well-learned movements remain exquisitely calibrated. A long-standing puzzle is how this process is applied when learning a new skill: Prior work has highlighted that implicit recalibration is inflexible, even when its operation results in maladaptive behavior given the sensorimotor mapping required for the new skill. Here, we revisited this problem, training participants for up to a month on a range of tasks with novel mappings, including mirror reversals and large rotations. Contrary to previous findings, implicit recalibration was flexibly reconfigured after prolonged training in a manner appropriate for each novel mapping. Training that emphasized feedforward planning rather than feedback control was sufficient to induce this reconfiguration. Learning involved reconfiguration of the sensory predictions and error correction policy for the new mapping, and participants were able to switch between mappings with minimal interference. These results demonstrate that implicit recalibration is reconfigured during skill acquisition, enabling flexible, context-specific adaptation.
Open AccessDevelopment of appropriate sensorimotor recalibration for a de novo skill
Implicit adaptation operates automatically to ensure that well-learned movements remain exquisitely calibrated. A long-standing puzzle is how this process is applied when learning a new skill: Prior work has highlighted that implicit recalibration is inflexible, even when its operation results in maladaptive behavior given the sensorimotor mapping required for the new skill. Here, we revisited this problem, training participants for up to a month on a range of tasks with novel mappings, including mirror reversals and large rotations. Contrary to previous findings, implicit recalibration was flexibly reconfigured after prolonged training in a manner appropriate for each novel mapping. Training that emphasized feedforward planning rather than feedback control was sufficient to induce this reconfiguration. Learning involved reconfiguration of the sensory predictions and error correction policy for the new mapping, and participants were able to switch between mappings with minimal interference. These results demonstrate that implicit recalibration is reconfigured during skill acquisition, enabling flexible, context-specific adaptation.
- Open Access
- 9 October 2026
- Published by Public Library of Science (PLoS) in PLOS Biology
- Vol. 24 (10) , e3004057
Implicit sensorimotor adaptation comprises distinct mechanisms for action selection and execution
The sensorimotor system is continuously adjusted to minimize error. Current theories assume that this adaptation process entails the operation of multiple learning systems, with a key division between implicit and explicit components. Recent studies have revealed several inconsistencies regarding the characteristics and constraints of the implicit system, suggesting that the current framework is incomplete. Here, we propose that these conflicting findings can be understood by recognizing that there are multiple implicit subcomponents, each with distinct computational goals. One well-studied component is implicit recalibration, a process critical for action execution which uses sensory-prediction errors to automatically refine the sensorimotor map. In the current study, we describe a second, novel component, implicit aiming, a process which contributes to action selection to achieve specific goals. Through a series of studies using human participants, we find compelling evidence that those two implicit processes show a clear separation in their temporal stabilities and contextual modulations. These distinct properties correspond to different computational frameworks attributing learning dynamics to either contextual inference or cancellation of competing neural populations, respectively. Together, these findings suggest an alternative framework for sensorimotor adaptation based on the computational goals of the system rather than phenomenology.
- 9 October 2026
- Published by Public Library of Science (PLoS) in PLOS Biology
- Vol. 24 (10) , e3004057
Open AccessImplicit sensorimotor adaptation comprises distinct mechanisms for action selection and execution
The sensorimotor system is continuously adjusted to minimize error. Current theories assume that this adaptation process entails the operation of multiple learning systems, with a key division between implicit and explicit components. Recent studies have revealed several inconsistencies regarding the characteristics and constraints of the implicit system, suggesting that the current framework is incomplete. Here, we propose that these conflicting findings can be understood by recognizing that there are multiple implicit subcomponents, each with distinct computational goals. One well-studied component is implicit recalibration, a process critical for action execution which uses sensory-prediction errors to automatically refine the sensorimotor map. In the current study, we describe a second, novel component, implicit aiming, a process which contributes to action selection to achieve specific goals. Through a series of studies using human participants, we find compelling evidence that those two implicit processes show a clear separation in their temporal stabilities and contextual modulations. These distinct properties correspond to different computational frameworks attributing learning dynamics to either contextual inference or cancellation of competing neural populations, respectively. Together, these findings suggest an alternative framework for sensorimotor adaptation based on the computational goals of the system rather than phenomenology.
- Open Access
Impaired representation of temporal statistics in cerebellar degeneration
To increase perceptual precision, the brain is hypothesized to perform Bayesian inference, evaluating sensory input within the context of the statistical structure of the environment. The cerebellum is a strong candidate for this computation in interval timing: its integrity is essential for precise timing and supports encoding temporal statistics in animals. However, how the sensory input is combined with prior knowledge within this circuit remains unknown. To examine this, we tested patients with cerebellar degeneration and matched controls on duration reproduction tasks with varying prior distributions. Patients showed the signatures of Bayesian inference, but with greater timing variability, stronger regression toward the mean, and slower updating of the prior. We developed a model of the cerebellar microcircuit in which Purkinje cell plasticity over a granule cell temporal basis set provides a shared substrate for learning the prior and performing the inference. Simulations of cerebellar degeneration showed an output approximated a Bayesian observer with an intact prior and a noisier likelihood. This model provided a good fit of the patients’ behavior over a wide range of parameter space. These results suggest a critical role for the human cerebellum in Bayesian inference, supporting optimally timed behavior even under extensive neuropathology.
Open AccessImpaired representation of temporal statistics in cerebellar degeneration
To increase perceptual precision, the brain is hypothesized to perform Bayesian inference, evaluating sensory input within the context of the statistical structure of the environment. The cerebellum is a strong candidate for this computation in interval timing: its integrity is essential for precise timing and supports encoding temporal statistics in animals. However, how the sensory input is combined with prior knowledge within this circuit remains unknown. To examine this, we tested patients with cerebellar degeneration and matched controls on duration reproduction tasks with varying prior distributions. Patients showed the signatures of Bayesian inference, but with greater timing variability, stronger regression toward the mean, and slower updating of the prior. We developed a model of the cerebellar microcircuit in which Purkinje cell plasticity over a granule cell temporal basis set provides a shared substrate for learning the prior and performing the inference. Simulations of cerebellar degeneration showed an output approximated a Bayesian observer with an intact prior and a noisier likelihood. This model provided a good fit of the patients’ behavior over a wide range of parameter space. These results suggest a critical role for the human cerebellum in Bayesian inference, supporting optimally timed behavior even under extensive neuropathology.
- Open Access
Statistical factorization in motor control: Opposing biases from statistics on different timescales enable fast and accurate behavior
Motor planning is fundamentally constrained by the speed–accuracy tradeoff: a single optimization that balances time and error costs, such that longer preparation yields more accurate movements while faster ones incur greater error. Here we propose that the motor system exploits environmental statistics to reshape this relationship, using two complementary computations that rely on distinct statistical information. The first leverages the marginal statistics of the target distribution, learned by integrating slowly over many trials, to modify the initial preparation state, accelerating planning for frequently encountered targets; this simultaneously speeds decisions and reduces planning variability, but introduces an attractive bias toward the mean of the target distribution. Strikingly, a second process tracks the sequential statistics of the environment over only the last few trials, producing a repulsive sequential bias whose persistence is tuned to the temporal autocorrelation and selectively counteracting the attractive bias introduced by the first. These results reveal that the motor system applies statistical factorization, estimating marginal and sequential statistics on their own timescales to shape movement preparation, thereby escaping the speed–accuracy tradeoff and producing movements that are both faster and more accurate.
Open AccessStatistical factorization in motor control: Opposing biases from statistics on different timescales enable fast and accurate behavior
Motor planning is fundamentally constrained by the speed–accuracy tradeoff: a single optimization that balances time and error costs, such that longer preparation yields more accurate movements while faster ones incur greater error. Here we propose that the motor system exploits environmental statistics to reshape this relationship, using two complementary computations that rely on distinct statistical information. The first leverages the marginal statistics of the target distribution, learned by integrating slowly over many trials, to modify the initial preparation state, accelerating planning for frequently encountered targets; this simultaneously speeds decisions and reduces planning variability, but introduces an attractive bias toward the mean of the target distribution. Strikingly, a second process tracks the sequential statistics of the environment over only the last few trials, producing a repulsive sequential bias whose persistence is tuned to the temporal autocorrelation and selectively counteracting the attractive bias introduced by the first. These results reveal that the motor system applies statistical factorization, estimating marginal and sequential statistics on their own timescales to shape movement preparation, thereby escaping the speed–accuracy tradeoff and producing movements that are both faster and more accurate.
- Open Access
- 19 February 2026
- Published by eLife Sciences Publications, Ltd in eLife
- Vol. 13
Motor biases reflect a misalignment between visual and proprioceptive reference frames
Goal-directed movements can fail due to errors in our perceptual and motor systems. While these errors may arise from random noise within these sources, they also reflect systematic motor biases that vary with the location of the target. The origin of these systematic biases remains controversial. Drawing on data from an extensive array of reaching tasks conducted over the past 30 years, we evaluated the merits of various computational models regarding the origin of motor biases. Contrary to previous theories, we show that motor biases produced by human participants do not arise from systematic errors associated with the sensed hand position during motor planning or from the biomechanical constraints imposed during motor execution. Rather, motor biases are primarily caused by a misalignment between eye-centric and body-centric representations of position. This model can account for motor biases across a wide range of contexts, encompassing movements with the right versus left hand, finger versus hand movements, visible and occluded starting positions, as well as before and after sensorimotor adaptation.
- 19 February 2026
- Published by eLife Sciences Publications, Ltd in eLife
- Vol. 13
Open AccessMotor biases reflect a misalignment between visual and proprioceptive reference frames
Goal-directed movements can fail due to errors in our perceptual and motor systems. While these errors may arise from random noise within these sources, they also reflect systematic motor biases that vary with the location of the target. The origin of these systematic biases remains controversial. Drawing on data from an extensive array of reaching tasks conducted over the past 30 years, we evaluated the merits of various computational models regarding the origin of motor biases. Contrary to previous theories, we show that motor biases produced by human participants do not arise from systematic errors associated with the sensed hand position during motor planning or from the biomechanical constraints imposed during motor execution. Rather, motor biases are primarily caused by a misalignment between eye-centric and body-centric representations of position. This model can account for motor biases across a wide range of contexts, encompassing movements with the right versus left hand, finger versus hand movements, visible and occluded starting positions, as well as before and after sensorimotor adaptation.
- Open Access
- 8 January 2026
- Published by eLife Sciences Publications, Ltd in eLife preprints
- Vol. 13
Motor biases reflect a misalignment between visual and proprioceptive reference frames
Goal-directed movements can fail due to errors in our perceptual and motor systems. While these errors may arise from random noise within these sources, they also reflect systematic motor biases that vary with the location of the target. The origin of these systematic biases remains controversial. Drawing on data from an extensive array of reaching tasks conducted over the past 30 years, we evaluated the merits of various computational models regarding the origin of motor biases. Contrary to previous theories, we show that motor biases do not arise from systematic errors associated with the sensed hand position during motor planning or from the biomechanical constraints imposed during motor execution. Rather, motor biases are primarily caused by a misalignment between eye-centric and body-centric representations of position. This model can account for motor biases across a wide range of contexts, encompassing movements with the right versus left hand, finger versus hand movements, visible and occluded starting positions, as well as before and after sensorimotor adaptation.
- 8 January 2026
- Published by eLife Sciences Publications, Ltd in eLife preprints
- Vol. 13
Open AccessMotor biases reflect a misalignment between visual and proprioceptive reference frames
Goal-directed movements can fail due to errors in our perceptual and motor systems. While these errors may arise from random noise within these sources, they also reflect systematic motor biases that vary with the location of the target. The origin of these systematic biases remains controversial. Drawing on data from an extensive array of reaching tasks conducted over the past 30 years, we evaluated the merits of various computational models regarding the origin of motor biases. Contrary to previous theories, we show that motor biases do not arise from systematic errors associated with the sensed hand position during motor planning or from the biomechanical constraints imposed during motor execution. Rather, motor biases are primarily caused by a misalignment between eye-centric and body-centric representations of position. This model can account for motor biases across a wide range of contexts, encompassing movements with the right versus left hand, finger versus hand movements, visible and occluded starting positions, as well as before and after sensorimotor adaptation.
- Open Access
- 31 October 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (44) , eadr4540
Contextual effects during sensorimotor adaptation are an emergent property of population coding in a cerebellar-inspired model
The cerebellum is crucial for sensorimotor adaptation, using error information to keep the sensorimotor system well calibrated. Here, we introduce a population-coding model to explore this process, focusing on how implicit adaptation is modulated by context. The model consists of a two-layered network, designed to capture activity in both the cerebellar cortex and deep cerebellar nuclei. A core feature of the model is that within each layer, the processing units are tuned to both movement direction and the direction of movement error. The model recapitulates a large range of contextual effects such as how prior learning, error uncertainty, or temporal delay affect the rate of adaptation. While these effects have traditionally been taken to indicate a form of meta-learning or context-dependent memory within the adaptation system, the current results show that they can emerge from the population activity of a cerebellar-like hierarchical network that does not represent context.
- 31 October 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (44) , eadr4540
Open AccessContextual effects during sensorimotor adaptation are an emergent property of population coding in a cerebellar-inspired model
The cerebellum is crucial for sensorimotor adaptation, using error information to keep the sensorimotor system well calibrated. Here, we introduce a population-coding model to explore this process, focusing on how implicit adaptation is modulated by context. The model consists of a two-layered network, designed to capture activity in both the cerebellar cortex and deep cerebellar nuclei. A core feature of the model is that within each layer, the processing units are tuned to both movement direction and the direction of movement error. The model recapitulates a large range of contextual effects such as how prior learning, error uncertainty, or temporal delay affect the rate of adaptation. While these effects have traditionally been taken to indicate a form of meta-learning or context-dependent memory within the adaptation system, the current results show that they can emerge from the population activity of a cerebellar-like hierarchical network that does not represent context.
- Open Access
Distinct Implicit Contributions to Action Selection and Action Execution in Sensorimotor Adaptation
The sensorimotor system is continuously adjusted to minimize error. Current theories assume that this adaptation process entails the operation of multiple learning systems, with a key division between implicit and explicit components. Recent studies have revealed several inconsistencies regarding the characteristics and constraints of the implicit system, suggesting that the current framework is incomplete. Here, we propose that these conflicting findings can be understood by recognizing that there are multiple implicit subcomponents, distinguished by their distinct computational goals. One well-studied component is implicit recalibration, a process critical for action execution which uses sensory-prediction errors to automatically refine the sensorimotor map. Here we describe a second, novel component, implicit aiming, a process which contributes to action selection to achieve the specific goals. Through a series of studies, we find compelling evidence that those two implicit processes show a clear separation in their temporal stabilities and contextual modulations. These distinct properties correspond to different computational frameworks attributing learning dynamics to either contextual inference or cancellation of competing neural populations, respectively. Together, these findings suggest an alternative framework for sensorimotor adaptation based on the computational goals of the system rather than phenomenology.
Open AccessDistinct Implicit Contributions to Action Selection and Action Execution in Sensorimotor Adaptation
The sensorimotor system is continuously adjusted to minimize error. Current theories assume that this adaptation process entails the operation of multiple learning systems, with a key division between implicit and explicit components. Recent studies have revealed several inconsistencies regarding the characteristics and constraints of the implicit system, suggesting that the current framework is incomplete. Here, we propose that these conflicting findings can be understood by recognizing that there are multiple implicit subcomponents, distinguished by their distinct computational goals. One well-studied component is implicit recalibration, a process critical for action execution which uses sensory-prediction errors to automatically refine the sensorimotor map. Here we describe a second, novel component, implicit aiming, a process which contributes to action selection to achieve the specific goals. Through a series of studies, we find compelling evidence that those two implicit processes show a clear separation in their temporal stabilities and contextual modulations. These distinct properties correspond to different computational frameworks attributing learning dynamics to either contextual inference or cancellation of competing neural populations, respectively. Together, these findings suggest an alternative framework for sensorimotor adaptation based on the computational goals of the system rather than phenomenology.
- Open Access
- 29 August 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (35) , eadv6572
Sequential effects in reaching reveal efficient coding in motor planning
The nervous system uses prior information to enhance movement accuracy, yet the underlying computational mechanisms remain relatively unclear. Prevailing motor control models emphasize Bayesian inference, where prior information is integrated to optimally estimate the current state. An alternative framework, efficient coding, proposes that the system dynamically reallocates encoding resources on the basis of environmental statistics—a mechanism highlighted in perception while underappreciated in motor control. We compared these frameworks in reaching movements, focusing on how the system leverages short-term priors in unpredictable environments. Unexpectedly, sequential effects aligned with the efficient coding model and contradicted Bayesian models. Specifically, current movements were biased in the opposite direction of previous movements, and movement variability decreased when successive reaches were similar. We further explored the temporal dynamics of these effects and showed that sequential bias is enhanced by intrinsic motor variability. These findings, accompanied by model comparisons, further support efficient coding in motor planning.
- 29 August 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (35) , eadv6572
Open AccessSequential effects in reaching reveal efficient coding in motor planning
The nervous system uses prior information to enhance movement accuracy, yet the underlying computational mechanisms remain relatively unclear. Prevailing motor control models emphasize Bayesian inference, where prior information is integrated to optimally estimate the current state. An alternative framework, efficient coding, proposes that the system dynamically reallocates encoding resources on the basis of environmental statistics—a mechanism highlighted in perception while underappreciated in motor control. We compared these frameworks in reaching movements, focusing on how the system leverages short-term priors in unpredictable environments. Unexpectedly, sequential effects aligned with the efficient coding model and contradicted Bayesian models. Specifically, current movements were biased in the opposite direction of previous movements, and movement variability decreased when successive reaches were similar. We further explored the temporal dynamics of these effects and showed that sequential bias is enhanced by intrinsic motor variability. These findings, accompanied by model comparisons, further support efficient coding in motor planning.