OSCR

Behavioral evidence for the hierarchical execution of sequential movements.

Code ↔ Paper

5 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 5 matches
  1. [1] § Methods › Simulation parameters ↔ paper_conf.py, lines 1–50 · score 0.62 · Lotka Volterra, paper conf, py, simulations, hierarchical
  2. [2] § Methods › Models › Via-point stochastic optimal control (vpSOC) ↔ single.py, lines 467–614 · score 0.62 · control matrix, optimal control, Kalman, Kt, filter, SOC
  3. [3] § Methods › Models › Dynamical system of an arm ↔ import_data.py, lines 45–136 · score 0.57 · low pass filter, arm model, motor, speed
  4. [4] § Methods › Simulation parameters ↔ paper_conf.py, lines 1–50 · score 0.56 · monitor_distance, monitor_sd, vpSOC, simulations, hierarchical, models
  5. [5] § Methods › Models › Via-point stochastic optimal control (vpSOC) ↔ single.py, lines 191–220 · score 0.52 · longer movements, weights, Todorov, noise, stochastic, optimal

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 289 lines · 11 KB · CC-BY-4.0 · 2 matches

  1. """Configuration parameters for the manuscript.
  3. The first section comprises general parameters for fonts and colors, followed
  4. by the sign-flipping rules for halfway and transitional coarticulations.
  6. The second section are the figue-specific parameters for simulations with vpSOC
  7. and HiSeq. Here we list only those parameters that are crucial for the
  8. figures. All others are left as the default parameters that can be found in
  9. each model:
  10. 1. The base class for SOC can be found in models/single.py: SOC. Its default
  11. parameters are:
  12. def_pars = {'sigma_c': 0.3,
  13. 'p_scaling': 1,
  14. 'm': 1.0,
  15. 'omega_nu': 0.2, # Adapted for longer movements
  16. 'omega_f': 0.3, # Adapted for longer movements
  17. 'N': 5,
  18. 'M': 3,
  19. 'Nc': 1,
  20. 'tau1': 0.04,
  21. 'tau2': 0.04,
  22. 'delta': 0.01,
  23. 'sigma_s': 0.5,
  24. 'r': 1e-5, # Weight of the energy term for cost
  25. 'obs_noises': [0.02, 0.02, 1],
  26. 'dt': 0.01,
  27. 't_end': 1,
  28. 'add_noise': 1000, # Additive noise for dynamics
  29. }
  30. 2. The class for vpSOC can be found in models/hierarchical.py:vpSOC. Besides
  31. parameters pertaining to the experimental task (targets, sequences, etc.), as
  32. well as those from SOC (using the same default values), it introduces
  33. h_scaling. h_scaling and r are systematically changed for simulations and
  34. defined differently for each figure.
  36. 3. The HiSeq class can be found in models/hierarchical.py:HiSeq. It uses the
  37. parameters from SOC and vpSOC, with the same defaults. In addition, it
  38. introduces a number of parameters for the dynamical system (Lotka-Volterra
  39. equations) which are left unchanged throughout simulations. It also introduces
  40. the monitor_distance parameter, which determines how close to the target (the
  41. center of the target circle) a trajectory has to get before the next element of
  42. the sequence is triggered. Related, the monitor_sd parameter controls the speed
  43. with which new elements are activated. These are varied throughout simulations.
  45. """
  46. from itertools import product
  48. import numpy as np
  50. from src.model import hierarchical as hie
  52. ####################################################
  53. # ------- General configuration for plots -------
  54. ####################################################
  57. # Data stuff
  58. all_parts = np.arange(401, 421)
  59. hiseq_parts = np.array([402, 403, 404, 418, 413])
  62. # Plotting stuff
  63. subplot_label_font = {'size': 12}
  65. colors_seq = np.array([[150, 150, 255], [150, 0, 255], [0, 150, 255],
  66. [255, 150, 150], [255, 150, 0], [255, 0, 150]]) / 256
  67. colors_sin = np.array([[0, 0, 255], [255, 0, 0]]) / 256
  70. # Analysis parameters
  72. # Depending on the function (trajectory_variance vs trajvar_all_segments),
  73. # different rules apply.
  74. # Flipping sign for trajectory_variance
  75. flip_rule_trans = {(0, 'sequential'): -1,
  76. (3, 'sequential'): -1,
  77. (5, 'sequential'): -1}
  78. flip_rule_half = {(1, 'sequential'): -1,
  79. (2, 'sequential'): -1,
  80. (4, 'sequential'): -1} # Flips crosses for readability
  82. # Flipping signs for trajvar_all_segments
  83. flip_rule_all = {(0, 'sequential', 0): -1,
  84. (4, 'sequential', 0): -1,
  85. (0, 'sequential', 1): -1,
  86. (1, 'sequential', 1): -1,
  87. (2, 'sequential', 1): -1,
  88. (3, 'sequential', 1): -1,
  89. (4, 'sequential', 1): -1,
  90. (5, 'sequential', 1): -1,
  91. (5, 'sequential', 2): -1}
  93. # Models parameters
  94. ref_vpsoc = {'soc_pars': {'add_noise': 10_000},
  95. 't_end': 0.7,
  96. 'h_scaling': 1
  97. }
  98. ref_sims = {}
  101. ####################################################
  102. # ------- For Figures 2 and 3, with data -------
  103. ####################################################
  104. good_part = 404 # Participant A
  105. bad_part = 406 # Participant B
  108. ####################################################
  109. # --------- For Figure 4 and 5, classification -----
  110. ####################################################
  112. # Selected participants for the classification of trajectories:
  113. parts_hiseq_class = [412, 414]
  116. # Model parameters for vpSOC(3)
  117. _r_values = [1e-8, 5e-8, 1e-7, 2e-7, 3e-7, 4e-7, 5e-7, 6e-7, 8e-7, 1e-6,
  118. 1.2e-6, 2e-6, 3e-6, 5e-6, 8e-6, 1e-5, 5e-5]
  119. # _r_values = [1e-8, 1e-7]
  120. _h1 = [0.6, 1, 1.3, 1.7, 2, 18]
  121. _h2 = [1, 2, 18]
  122. _h3 = [5]
  123. vpsoc3_pars = {'r': _r_values,
  124. 'h': list(product(_h1, _h2, _h3))}
  127. # Model parameters for HiSeq simulations
  128. dd = 10 # Dummy value; unused by models but may be expected
  129. h_sec = 10
  130. hs = [1.1, 1.2, 1.5, 1.8]
  131. all_h = [np.array([[dd, dd], *[[h_prot, 20, 0]] * 6, *[[h_sec, dd]] * 3],
  132. dtype=object)
  133. for h_prot in hs]
  134. all_r = [1.0e-07, 5.0e-07, 7.0e-07, 9.0e-07, 1.0e-06, 1.2e-06, 1.5e-06,
  135. 1.8e-06, 2.0e-06, 2.2e-06, 2.5e-06, 3.0e-06, 4.0e-06, 5.0e-06,
  136. 6.0e-06, 7.0e-06, 8.0e-06, 1.0e-05, 2.0e-05, 3.0e-05, 5.0e-05,
  137. 1.0e-04, 5.0e-04, 1.0e-03]
  138. all_r = all_r[-2:]
  139. all_mondis = [0.1, 0.4, 0.5, 0.7, 1, 1.5, 2]
  140. # all_mondis = [0.05, 0.5, 2]
  141. hiseq_total = len(all_h) * len(all_r) * len(all_mondis)
  142. hiseq_parprod = enumerate(product(all_h, all_r, all_mondis))
  145. # # Other (fixed) Model parameters for HiSeq simulations
  146. # tau23 = 3
  147. # monitor_sd = 10
  148. # monitor_distance = 1.2
  149. # mondis_abs = False
  150. # transitions = [[0, 0], [0, 1], [1, 3], [3, 6], # all SOC
  151. # [0, 1, 3], # vpSOC(2) + SOC
  152. # [1, 3, 6]] # SOC + vpSOC(2)
  154. # # All seqs. are idmove 0, with three "models", repeated to make 6:
  155. # sequences = np.array([[0, 1, 2, 3], [0, 4, 3], [0, 1, 5]] * 2, dtype=object)
  156. # t_ends = 1.7 # [1, 1, 1, 1, 2, 2]
  157. # r = [1e-4]
  158. # soc_pars = {'r': r, 'add_noise': 10}
  159. # h_scalings = []
  160. # for ch1 in [10, 30, 50]:
  161. # h_scalings.append([[1], [1], [1], [1], [ch1, 1], [ch1, _h3[0]]])
  162. # hiseq_pars = {'tau23': tau23, 'monitor_sd': monitor_sd,
  163. # 'monitor_distance': monitor_distance, 'transitions': transitions,
  164. # 'sequences': sequences, 't_ends': t_ends, 'r': r,
  165. # 'h': h_scalings, 'soc_pars': soc_pars,
  166. # 'mondis_abs': mondis_abs}
  169. ####################################################
  170. ####################################################
  171. ############# Supplementary figures ################
  172. ####################################################
  173. ####################################################
  175. # Note that some of these are no longer used.
  177. ####################################################
  178. # --------- For Supp. Figure, cost of planning------
  179. ####################################################
  181. cost_num_targets = np.arange(1, 5)
  182. cost_durations = np.arange(0.5, 20, 0.5)
  183. cost_reps = 1000
  186. ####################################################
  187. # ------- For Figure 5, with vpSOC simulations -----
  188. ####################################################
  189. r_low_high = [4e-6, 8e-6]
  191. # For vpSOC(3):
  192. hs_first = np.array([0.9, 10])
  193. hs_second = np.array([1.5, 18])
  194. hs_third = np.array([15])
  195. hs_vpsoc3 = list(product(hs_first, hs_second, hs_third))
  197. # For vpSOC(2):
  198. hs_first = np.array([0.4, 3])
  199. hs_second = np.array([3])
  200. hs_vpsoc2 = list(product(hs_first, hs_second))
  203. ####################################################
  204. # ------- For Figure 8, with data and models -------
  205. ####################################################
  207. # Models and parameters
  208. hiseq_ts = np.array([1, 2, 1]) * 1
  209. tau23 = 3
  210. monitor_sd = 5
  211. monitor_distance = 30
  212. r_high = 2e-5
  213. r_low = 2e-7
  214. mods = {'vpSOC(3)+': (hie.vpSOC,
  215. [{'h_scaling': [15, 15, 6],
  216. 't_ends': 1.6,
  217. 'soc_pars': {'add_noise': 0,
  218. 'r': r_high}},
  219. {'h_scaling': [6, 10, 15],
  220. 't_ends': 1.6,
  221. 'soc_pars': {'add_noise': 0,
  222. 'r': r_high}},
  223. {'h_scaling': [2, 8, 15],
  224. 't_ends': 1.6,
  225. 'soc_pars': {'add_noise': 0,
  226. 'r': r_high}},
  227. ]
  228. ),
  229. 'vpSOC(3)-': [hie.vpSOC,
  230. [{'h_scaling': [1, 7, 15],
  231. 't_ends': 1.6,
  232. 'soc_pars': {'add_noise': 8_000,
  233. 'r': r_low}
  234. },
  235. {'h_scaling': [1, 7, 15],
  236. 't_ends': 1.6,
  237. 'soc_pars': {'add_noise': 0,
  238. 'r': r_low}
  239. },
  240. {'h_scaling': [1, 7, 15],
  241. 't_ends': 1.6,
  242. 'soc_pars': {'add_noise': 0,
  243. 'r': r_low}
  244. },
  245. ]
  246. ],
  247. 'HiSeq (vpSOC(2)+vpSOC(1))': [hie.HiSeq,
  248. [{'transitions': [[0, 0], [0, 1, 3], [3, 6]],
  249. 'sequences': [[0, 1, 2]] * 6,
  250. 't_ends': hiseq_ts, # TODO: fix this
  251. 'h_scaling': 5,
  252. 'tau23': tau23,
  253. 'monitor_sd': monitor_sd,
  254. 'monitor_distance': monitor_distance,
  255. 'soc_pars': {'add_noise': 90,
  256. 'r': 5e-5}},
  257. {'transitions': [[0, 0], [0, 1, 4], [4, 6]],
  258. 'sequences': [[0, 1, 2]] * 6,
  259. 't_ends': hiseq_ts, # TODO: fix this
  260. 'h_scaling': np.array([[1, 1], [15, 3],
  261. [2, 1]]),
  262. 'tau23': tau23,
  263. 'monitor_sd': monitor_sd,
  264. 'monitor_distance': monitor_distance,
  265. 'soc_pars': {'add_noise': 90,
  266. 'r': 5e-5}},
  267. {'transitions': [[0, 0], [0, 2, 3], [3, 6]],
  268. 'sequences': [[0, 1, 2]] * 6,
  269. 't_ends': hiseq_ts, # TODO: fix this
  270. 'h_scaling': np.array([10, 5]),
  271. 'tau23': tau23,
  272. 'monitor_sd': monitor_sd,
  273. 'monitor_distance': monitor_distance,
  274. 'soc_pars': {'add_noise': 90,
  275. 'r': 5e-5}},
  276. ]
  277. ]
  278. }
  279. # Selected participant data (part, id_move)
  280. example_trajs = {'vpSOC(3)+': [(406, 0),
  281. (414, 2),
  282. (420, 3)],
  283. 'vpSOC(3)-': [(413, 0),
  284. (413, 2),
  285. (413, 3)],
  286. 'HiSeq (vpSOC(2)+vpSOC(1))': [(418, 0),
  287. (403, 2),
  288. (412, 3)]
  289. }

paper_conf.py, under CC-BY-4.0 · at the source

Overview

Authors: Darío Cuevas Rivera1,2, Stefan J. Kiebel1,2
ORCID iDs: Darío Cuevas Rivera, Stefan J. Kiebel
  1. Faculty of Psychology, Technische Universität Dresden,Dresden, Germany
  2. Centre for Tactile Internet with Human-in-the-Loop (CeTI),Dresden, Germany
Institutions: Technische Universität Dresden (Germany)
Journal: Communications psychology, volume 4, issue 1, article 52
Dates: received 16 June 2025; accepted 26 February 2026; published online 11 March 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1038/s44271-026-00436-5 · PMID 41807720 · PMCID PMC13009195 · OpenAlex W7134979970
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: computational modeling (no new data) (modality), human (organism)
Methods: Statistics, Smoothing, state filtering, decompositions
Keywords: Computational models, Motor cortex, Human behaviour
Topic: Motor Control and Adaptation (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Citations: cited by 1 paper (Europe PMC); 54 references in the paper

Abstract

Movements in humans and other animals are known to be hierarchically organized, with simple movements forming the building blocks to more complex, sequential movements, a phenomenon often referred to as chunking. Neuroimaging studies have highlighted this layered structure, implicating the primary motor cortex in simple movements, and pre-motor and parietal areas in sequences of movements. Behavioral experiments designed to study this hierarchy have required extensive training of simple movement sequences, such as key presses, using error rates and reaction times as indirect markers of chunking. In this work, we provide kinematic evidence that the hierarchical organization of movements naturally emerges during reaching movements toward large targets, without the need for extensive training. Through model-based analyses of the observed trajectories’ geometry in a sequential pointing task (N = 20 participants), we infer the underlying hierarchy of the mechanism guiding movement generation. Our results show that most participants adapt their strategy dynamically using hierarchical planning, depending on the sequence. These findings offer insights into the process of chunking, as well as the conditions on how and when humans switch between flat and hierarchical planning during movement.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 5 matches between paragraphs and lines of code.

figshare 31169632

License: CC-BY-4.0
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Languages: Python (22)
Size: 49 files, 22 scripts
Software Heritage: not checked
Found in: the references
Holds: README, environment (poetry.lock, pyproject.toml)
Not found: license file, CITATION.cff, tests, continuous integration, documentation
Tools: NumPy (17 files), pandas (9 files), Matplotlib (6 files), SciPy (4 files), seaborn (3 files), PsychoPy (2 files)
Availability: 1 check, the latest on 30 September 2026: the link answers (HTTP 200)
  • 30 September 2026: the link answers (HTTP 200)
22 files
At the source:

Code availability

The code used to generate the figures and results in this work can be found in an online repository (link to repository (https://www.doi.org/10.6084/m9.figshare.31169632))54. Instructions to set up a Python environment to use the code are contained within the README.md file. The pre-processed behavioral data and simulated data used for the results can also be found in this repository.

Reproduced under the paper's license (CC BY), from the paper cited above.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 21 scripts, each with its path and the digest of its content;
  • 5 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

Datasets cited

Data availability

The raw behavioral data were published in an online repository (link to repository (https://doi.org/10.6084/m9.figshare.31188148.v1))52 in the Motion-BIDS standard53.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 3 keywords, 53 references.

Cite

This paper

Cuevas Rivera, D., & Kiebel, S. J. (2026). Behavioral evidence for the hierarchical execution of sequential movements. Communications psychology, 4(1), 52. https://doi.org/10.1038/s44271-026-00436-5

BibTeX

@article{cuevasrivera2026behavioral,
author = {Cuevas Rivera, Darío and Kiebel, Stefan J.},
title = {{Behavioral evidence for the hierarchical execution of sequential movements}},
journal = {Communications psychology},
year = {2026},
month = mar,
volume = {4},
number = {1},
pages = {52},
publisher = {Nature Publishing Group},
issn = {2731-9121},
doi = {10.1038/s44271-026-00436-5},
url = {https://doi.org/10.1038/s44271-026-00436-5},
pmid = {41807720},
pmcid = {PMC13009195}
}

RIS

TY - JOUR
AU - Cuevas Rivera, Darío
AU - Kiebel, Stefan J.
TI - Behavioral evidence for the hierarchical execution of sequential movements
T2 - Communications psychology
J2 - Commun Psychol
PY - 2026
DA - 2026/03/11
VL - 4
IS - 1
SP - 52
SN - 2731-9121
PB - Nature Publishing Group
DO - 10.1038/s44271-026-00436-5
UR - https://doi.org/10.1038/s44271-026-00436-5
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s44271-026-00436-5",
"type": "article-journal",
"title": "Behavioral evidence for the hierarchical execution of sequential movements",
"container-title": "Communications psychology",
"author": [
{
"family": "Cuevas Rivera",
"given": "Darío"
},
{
"family": "Kiebel",
"given": "Stefan J."
}
],
"container-title-short": "Commun Psychol",
"volume": "4",
"issue": "1",
"page": "52",
"DOI": "10.1038/s44271-026-00436-5",
"PMID": "41807720",
"PMCID": "PMC13009195",
"ISSN": "2731-9121",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s44271-026-00436-5",
"language": "en",
"issued": {
"date-parts": [
[
2026,
3,
11
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1002/hbm.70562 [code]
Distinct Physiological Mechanisms Drive Grey Matter Plasticity in Complex Versus Simple Sequence Learning.
Journal: Human brain mapping
In common: seaborn, pandas, SciPy, 2 other tools, 5 references
[2] doi:10.1167/jov.26.8.4 [code]
The neural processes of illusory occlusion in object recognition.
Journal: Journal of vision
In common: PsychoPy, seaborn, pandas, 3 other tools, 1 reference
[3] doi:10.1038/s41598-026-53525-0 [code]
Timing-induced illusory percepts of pitch.
Journal: Scientific reports
In common: PsychoPy, seaborn, pandas, 3 other tools, 1 reference
[4] doi:10.34133/csbj.0042 [code]
Using Steady-State Visual Evoked Potentials to Characterize Wide-Ranging Retinopathy Linked to <i>CRB1</i>: Implications for Clinical Trials.
Journal: Computational and structural biotechnology journal
In common: PsychoPy, seaborn, pandas, 3 other tools, 1 reference
[5] doi:10.1162/nol.a.271 [code]
Compositional Complexity in Text and Images.
Journal: Neurobiology of language (Cambridge, Mass.)
In common: PsychoPy, seaborn, pandas, 3 other tools, 1 reference
[6] doi:10.3390/biomimetics11080569 [code]
Pretraining of Embodied Recurrent Networks Bridges the Gap Between Artificial and Cortical Neural Activities.
Journal: Biomimetics (Basel, Switzerland)
In common: SciPy, Matplotlib, NumPy, 3 references
[7] doi:10.7554/elife.111876 [code]
Distinct sensorimotor encoding in tuft dendrites and somata associated with action, correction, and learning.
Journal: eLife
In common: seaborn, pandas, SciPy, 2 other tools, 2 references
[8] doi:10.1016/j.dcn.2026.101791 [code]
Age-related differences in hippocampal network engagement during safety processing in adolescents.
Journal: Developmental cognitive neuroscience
In common: PsychoPy, seaborn, pandas, 3 other tools
[9] doi:10.1038/s41467-026-73865-9 [code]
Histamine shapes the neurocomputational dynamics of human learning.
Journal: Nature communications
In common: PsychoPy, seaborn, pandas, 3 other tools
[10] doi:10.1038/s41598-026-50946-9 [code]
Fixation-related potentials reveal that confusing program code elicits a late frontal positivity.
Journal: Scientific reports
In common: PsychoPy, seaborn, pandas, 3 other tools

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.