Pith. sign in

REVIEW 1 major objections 5 minor 50 references

A graph-native Python reimplementation of the IDyOM music-expectation model matches the Lisp reference to within 0.003 bits of information content while exposing its learned memory structures as inspectable graphs.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-01 01:27 UTC pith:RPNZJJP6

load-bearing objection Solid, honestly-scoped reimplementation with strong validation for the covered configurations; the 'faithful' claim needs a little more evidence for the untested viewpoints. the 1 major comments →

arxiv 2607.25787 v1 pith:RPNZJJP6 submitted 2026-07-28 cs.SD q-bio.NC

GraphIDyOM: A graph-native Python reimplementation of IDyOM for musical expectation modelling

Lluc Bono Rossell\'o This is my paper
classification cs.SD q-bio.NC
keywords musical expectationIDyOMinformation contententropyPrediction by Partial Matchinggraph representationreimplementationrecency-sensitive memory
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper claims that the influential IDyOM model of musical expectation can be reimplemented in Python with its long-term and short-term predictive memories represented as explicit directed graphs, without losing numerical fidelity. Against the original Lisp implementation, on 185 Bach chorale melodies and across direct, projected, multi-joint, and multi-source viewpoints, event-wise information-content differences stay below 0.003 bits and correlations stay above r=.9997. The point matters because IDyOM's internal memory has been a black box: making it a graph object lets researchers inspect, export, modify, and query the exact evidence behind each prediction. The paper then shows three uses this enables: network analysis of expectation topology, recency-sensitive memory retrieval, and a local server for interactive or experimental control of predictions.

Core claim

The paper's central claim is that the count-based variable-order prediction machinery of IDyOM—multiple viewpoints, source-to-target projection, PPM escape and exclusion handling, and entropy-weighted LTM/STM merging—can be exactly preserved while representing each Markov order's context–continuation statistics as a directed graph with weighted edges. Validated against Lisp IDyOM under matched settings, the implementation yields mean absolute information-content differences below 0.003 bits and Pearson correlations of at least 0.9997 across all evaluated configurations, including projected and multiple-viewpoint cases. The graph is not a compression or approximation; it is the memory itself,

What carries the argument

Order-specific n-gram memory graphs: for each Markov order, nodes are context windows and weighted directed edges record observed continuations, so prediction is a lookup from the current context node's outgoing edges. Around these graphs sits the IDyOM pipeline: viewpoint encoding and projection, PPM variable-order combination across orders within each memory (escape methods a, b, c, d, x; exclusion and update-exclusion), and final LTM/STM merge via entropy-weighted geometric combination with a confidence bias of 7.0 in the validated configuration. The work of the graph representation is to make counts and contexts inspectable and modifiable as first-class objects.

Load-bearing premise

The parity numbers presume that GraphIDyOM's parser and viewpoint encoders are exactly equivalent to Lisp IDyOM for the evaluated settings; the claim that the full 'faithful reimplementation' extends to the whole Lisp viewpoint ecosystem depends on the assumption that the unimplemented metrical, harmonic, dynamic, and specialised viewpoints would only add encoding and projection logic and would not disturb the validated prediction core.

What would settle it

Run the identical Table 1 validation protocol (185 Bach chorales, five-fold, seed 0, maximum order 5, LTM escape method c, STM escape method x, geometric LTM/STM merge with bias 7.0 and no offset) with a different random seed, say seed 1, and check all twelve rows: if any row shows a mean absolute information-content difference above 0.003 bits or a correlation below 0.9997, the parity claim is refuted. Alternatively, port an unimplemented Lisp viewpoint such as a metrical or harmonic viewpoint into GraphIDyOM and observe whether event-wise correlation drops, which would falsify the additive-e

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Researchers can now reproduce Lisp IDyOM's numerical outputs in pure Python, removing the Lisp dependency for the evaluated configurations.
  • Because memories are graphs, studies can map surprise and uncertainty onto musical topology, asking whether unexpected events sit on hubs, bridges, or peripheral states.
  • The optional recency-sensitive retrieval shows that stored context can be reweighted without retraining, enabling tests of forgetting and memory-decay hypotheses within the same architecture.
  • The local server turns expectation into a queryable, real-time signal, allowing stimuli to be designed to target entropy and information-content profiles, and interactive systems to use probabilities as control signals.
  • The framework's coverage of direct, projected, joint, and multi-source viewpoints exceeds that of previously available Python baselines while running with millisecond-level prediction latency.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the parity claim holds broadly, GraphIDyOM effectively makes the Lisp implementation's behaviour auditable: any future discrepancy can be traced to a specific node or edge in the memory graph, a transparency advantage the Lisp system does not provide.
  • The graph-native memory could serve as a substrate for differential testing: deliberately mutating edge weights, such as removing low-count edges, and observing downstream information-content changes would allow causal analysis of which corpus statistics drive particular predictions.
  • The paper's 'faithful' label currently rests on a set of 12 configurations over one corpus; if untested viewpoints such as harmonic or metrical ones require changes beyond additive encoders, the extrapolation would fail. A natural test would be to port one additional viewpoint family and re-check parity.
  • The interactive server architecture suggests a broader class of applications: expectation-aware generative music systems could use next-event probabilities as interpretable constraints, which the graph representation also makes editable per context.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

1 major / 5 minor

Summary. The paper introduces GraphIDyOM, a Python reimplementation of the IDyOM model for musical expectation. The implementation represents long-term and short-term predictive memories as explicit directed graphs at each Markov order, while retaining IDyOM's variable-order PPM prediction machinery, viewpoint projection, and LTM/STM merging. The authors validate the system against the original Lisp IDyOM on 185 Bach chorales using five-fold cross-validation: across 12 direct, projected, multi-joint, and multi-source configurations, mean absolute information-content differences are below 0.003 bits and Pearson correlations with the Lisp traces are at least r = .9997 (Section 3, Table 1). They also benchmark computational performance against Lisp IDyOM and IDyOMpy, and demonstrate extensions enabled by the graph-native representation: network analysis of learned memories, expectation-annotated musical networks, recency-sensitive retrieval based on PPM-decay, and a local server interface for interactive applications. The paper claims that GraphIDyOM is a faithful and accessible reimplementation that exposes internal memory structures for analysis and modification, while acknowledging that the full Lisp viewpoint ecosystem and automated viewpoint selection are not yet implemented.

Significance. If the claims hold, GraphIDyOM would be a valuable contribution to computational music cognition: it provides an open-source Python implementation with exceptionally strong numerical parity to the reference Lisp IDyOM on the tested configurations, and it makes the predictive memory inspectable and modifiable, which is a genuine methodological advance over black-box wrappers. The validation design is a clear strength: the implementation is compared against an external reference (Lisp IDyOM), the test configurations cover several architecturally distinct prediction modes, and the free parameters (escape methods, merge bias, max order) are inherited from prior IDyOM literature rather than fitted to the test set. The open-source code and reproducible validation protocol further strengthen the paper. The main reservation is that the broad 'faithful reimplementation' claim is supported only for a subset of the reference architecture's viewpoint ecosystem; the extrapolation to the full ecosystem is plausible but untested.

major comments (1)
  1. [§5 (Limitations) and Abstract/Introduction] The central claim of a 'faithful reimplementation' is broader than what the validation supports. Table 1 covers 12 configurations built from pitch, interval, contour, duration, and timing viewpoints on monophonic chorales, but Section 5 explicitly states that metrical, harmonic, dynamic, and specialised viewpoints, as well as automated viewpoint selection, are not implemented. The sentence 'extending this coverage should primarily require additional viewpoint encoding and projection logic rather than changes to the validated prediction core' is an extrapolation, not a demonstrated property. If any omitted viewpoint changes the core PPM/merge path — for example, a viewpoint with non-atomic symbols, a different target alphabet, or an interaction with exclusion/update-exclusion — the validated core may not suffice. This is load-bearing because the abstract and introduction claim a faithful
minor comments (5)
  1. [§3, Table 1] The validation statistics are reported as aggregate point estimates over the full held-out set: mean absolute ΔIC and Pearson r. There is no per-fold dispersion, standard deviation, or confidence interval. Given the very small differences this is unlikely to change the qualitative conclusion, but reporting fold-level ranges would make the parity claim more robust and would allow readers to assess variability across training splits.
  2. [§3, IDyOMpy comparison] The comparison with IDyOMpy is only approximate for several rows, as the paper acknowledges. For 'Pitch+octave' the IDyOMpy row uses 'pitch' rather than 'cpitch', and the table note says 'length-like rows use the closest aligned subset where required.' This makes the large ΔIC values for IDyOMpy not strictly comparable to the GraphIDyOM values. Please state explicitly which events and settings were used in each IDyOMpy row, or soften the comparison to avoid overstating the contrast.
  3. [§4.2, Figure 6] The recency-sensitive retrieval demonstration applies PPM-decay to the graph counts but does not compare the resulting traces to the published PPM-Decay model of Harrison et al. (2020) or to any reference implementation. The paper already labels this as a methodological demonstration rather than a validation, which is appropriate; however, if the framework is advertised as supporting recency-sensitive PPM, a numerical check against the reference would strengthen the claim.
  4. [§5, automated viewpoint selection] The absence of automated viewpoint selection is more than a missing convenience: in Lisp IDyOM, viewpoint selection is part of the modelling workflow and affects which predictions are made for a given dataset. This limitation should be stated earlier in the paper (e.g., in the introduction or Section 2) so that readers are not surprised when the validation and examples use manually specified viewpoints.
  5. [§3, Figure 2] The text says the GraphIDyOM traces are 'visually indistinguishable' from Lisp IDyOM. This is a qualitative statement; the quantitative Table 1 already makes the point. Consider reporting the maximum absolute difference for the illustrated chorale, or at least noting that the figure is illustrative of the aggregate result.

Circularity Check

0 steps flagged

No significant circularity: the core reimplementation claim is validated against the external Lisp IDyOM reference, and the only self-citations are in non-load-bearing demonstrations.

full rationale

The central claim of GraphIDyOM is that it faithfully reimplements IDyOM, and this is supported by direct numerical comparison against the original Lisp IDyOM, an external reference system, not against GraphIDyOM's own outputs. The paper reports that 'mean absolute information-content differences remained below 0.003 bits and correlations with the Lisp traces were at least r=.9997' across matched parser, viewpoint, PPM, update-exclusion, and LTM/STM merge settings. The free parameters (merge bias 7.0, escape methods c/x, max order 5) are inherited from prior IDyOM work, not fitted in this paper, and the validation corpus is the same external Bach chorale corpus used in the original IDyOM evaluation. There is no equation in the paper that reduces a predicted quantity to a fitted input; the formulas for entropy, PPM combination, and entropy-weighted merging are standard and independently sourced. The extension demonstrations (network analysis, PPM-decay, IRIDyOM) are presented as methodological examples and do not feed back into the validation. Section 5 concedes that metrical, harmonic, dynamic, and specialised viewpoints plus automated viewpoint selection are not implemented; the statement that extending coverage 'should primarily require additional viewpoint encoding and projection logic rather than changes to the validated prediction core' is an extrapolation and a limitation, not a circular derivation. The paper does cite the author's own prior work (references 33, 41, 48), but these appear in background and application-demonstration contexts, not as load-bearing evidence for the faithful-reimplementation claim. Accordingly, no specific circular step can be exhibited, and the appropriate score is at the low end of the no-significant-circularity range.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

All free parameters/custom choices are either inherited from the reference IDyOM configuration (max order 5, escape methods c/x, merge bias 7.0, δ=0) or left unspecified in the paper (PPM-decay buffer/decay). No new physical or mathematical entities are introduced; the graph objects are a representational choice, not a postulated mechanism. The main background assumptions are that Lisp IDyOM is a correct reference, that the Bach corpus is equivalently parsed, and that the graph encoding does not alter count-based PPM semantics.

free parameters (5)
  • Maximum Markov order K = 5 (validation; up to 15 in benchmark)
    Context depth used in validation; chosen from the reference IDyOM setup, not fitted by this paper.
  • LTM/STM escape methods = LTM: c, STM: x
    PPM escape-method choices in Section 3; predictions are sensitive to this setting; taken from the reference configuration.
  • LTM/STM merge bias b = 7.0
    Exponent in entropy-weighted combination Eq. 4; set to reference IDyOM value, not fitted here.
  • Entropy-weight offset δ = 0
    Offset in Eq. 4; disabled ('no entropy-weight offset') in the validation configuration.
  • PPM-decay buffer/decay parameters = not stated
    Used for the recency-sensitive retrieval demonstration in Section 4.2; values are not reported, so the demonstration is not reproducible as specified.
axioms (5)
  • domain assumption Lisp IDyOM is treated as ground truth for validation.
    Section 3 compares GraphIDyOM against Lisp as the reference; if the reference has version-specific behavior, GraphIDyOM inherits it rather than correcting it.
  • domain assumption The Bach chorale corpus from Pearce (2005) is equivalent to the data used by the original IDyOM and IDyOMpy, with matched preprocessing.
    The validation requires that all implementations see the same symbolic event stream; the paper states this but does not ship the corpus.
  • domain assumption PPM escape mechanisms and entropy-weighted geometric merge (Eqs. 3-5, from Cleary & Witten 1984 and Pearce et al. 2004) are implemented faithfully.
    The paper provides no independent derivation; it relies on the cited formulations and on its own implementation care.
  • domain assumption The directed n-gram graph representation preserves count-based PPM semantics exactly, including tie-breaking and iteration order.
    Central to the graph-native claim; not proven formally, only supported by the empirical agreement with Lisp.
  • domain assumption The five-fold seed-0 split and selected configurations are representative enough to generalize the agreement claim.
    The validation covers 9,227 events in eleven configurations but not the full viewpoint ecosystem; the paper's own limitations section acknowledges this.

pith-pipeline@v1.3.0-alltime-deepseek · 13750 in / 16874 out tokens · 162314 ms · 2026-08-01T01:27:48.138461+00:00 · methodology

0 comments
read the original abstract

The Information Dynamics of Music model (IDyOM) has played a central role in computational accounts of musical expectation by providing event-by-event estimates of uncertainty and surprise from symbolic musical sequences. However, its reference implementation is difficult to integrate with contemporary Python workflows, and its internal memory structures are not easily accessible for inspection or modification. We introduce GraphIDyOM, a graph-native Python reimplementation of IDyOM that represents long-term and short-term predictive memories as explicit graph objects while preserving the model's variable-order, multiple-viewpoint architecture. GraphIDyOM returns event-wise information content and entropy, exposes internal memory structures for analysis and export, and supports access through a local server. We validate the implementation against the original Lisp IDyOM across single, projected, and multiple-viewpoint configurations, and benchmark its coverage and computational performance against a recent reimplementation. We then demonstrate how the explicit memory representation supports network analysis of learned memories, projection of expectation values onto musical networks, recency-sensitive memory retrieval, and interactive applications. GraphIDyOM therefore provides both a faithful and accessible reimplementation of a widely used model and a platform for studying musical expectation through memory, topology, and interaction.

Figures

Figures reproduced from arXiv: 2607.25787 by Lluc Bono Rossell\'o.

Figure 1
Figure 1. Figure 1: GraphIDyOM schematic. Symbolic input is queried against order-specific long-term (LTM) and short-term (STM) memory graphs. At each order, the current context is matched to a node and outgoing weighted edges define possible continuations. PPM combines evidence across progressively shorter contexts within each memory, after which the LTM and STM distributions are merged to produce the final next-event distri… view at source ↗
Figure 2
Figure 2. Figure 2: illustrates the comparison for an individual held-out chorale under direct pitch-plus-octave prediction. The GraphIDyOM information-content and entropy traces are visually indistinguishable from the Lisp reference, whereas IDyOMpy shows larger deviations under the closest available configuration [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Pitch-octave computational performance benchmark. Computational profile for direct pitch-plus-octave prediction on the Bach dataset. The left column varies the number of LTM training files while keeping maximum Markov order fixed at 5. The right column fixes the LTM at 150 training files and varies maximum order. The top row shows LTM construction time; the bottom row shows average held-out prediction time… view at source ↗
Figure 4
Figure 4. Figure 4: Example network analysis of GraphIDyOM memories. Pitch-plus-octave memory graphs for one held-out Bach chorale. The left column shows LTM layers trained from 150 other Bach files, and the right column shows STM layers accumulated while processing the held-out piece. Rows show Markov orders 3, 2, and 1. Orange nodes and edges indicate the trajectory of the held-out piece through the corresponding LTM layer.… view at source ↗
Figure 5
Figure 5. Figure 5: shows this representation for a held-out Bach chorale. Panel A describes the piece using its own first-order pitch-plus-octave transition statistics. Panel B retains the same topology but annotates it with average predictions from an LTM-only model trained on other Bach chorales. The LTM-only condition is used to isolate expectations derived from prior exposure from statistics learned online from the displ… view at source ↗
Figure 6
Figure 6. Figure 6: illustrates the effect of recency-sensitive retrieval on a held-out Bach chorale. Panels A and B show event-wise predictions from the full LTM+STM model, while Panels C and D show the final order-1 STM graph accumulated over the same piece. The two conditions use the same pitch-plus-octave configuration and LTM training set and differ only in whether stored observations contribute through empirical counts … view at source ↗
Figure 7
Figure 7. Figure 7: Interactive deployment of GraphIDyOM. Assistant mode of the IRIDyOM Max for Live interface integrated in Ableton Live. GraphIDyOM provides the predictive backend, returning likely next-note candidates and event-wise information-content estimates while the musical sequence is inspected or edited. Adapted from Bono Rosselló et al. [41]. 13 [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 16 canonical work pages

  1. [1]

    Saffran, Elizabeth K

    Jenny R. Saffran, Elizabeth K. Johnson, Richard N. Aslin, and Elissa L. Newport. Statistical learning of tone sequences by human infants and adults.Cognition, 70(1):27–52, 1999. doi: 10.1016/S0010-0277(98)00075-4

    work page doi:10.1016/s0010-0277(98)00075-4 1999

  2. [2]

    MIT Press, 2006

    David Huron.Sweet Anticipation: Music and the Psychology of Expectation. MIT Press, 2006

    2006

  3. [3]

    MIT Press, Cambridge, MA, 2007

    David Temperley.Music and Probability. MIT Press, Cambridge, MA, 2007

    2007

  4. [4]

    Pearce and Geraint A

    Marcus T. Pearce and Geraint A. Wiggins. Auditory expectation: the information dynamics of music perception and cognition.Topics in Cognitive Science, 4(4):625–652, 2012

    2012

  5. [5]

    Marcus T. Pearce. Statistical learning and probabilistic prediction in music cognition: mechanisms of stylistic enculturation.Annals of the New York Academy of Sciences, 1423(1):378–395, 2018

    2018

  6. [6]

    Darrell Conklin and Ian H. Witten. Multiple viewpoint systems for music prediction.Journal of New Music Research, 24(1):51–73, 1995

    1995

  7. [7]

    IEEE Transactions on Communications, 32(4):396–402, 1984

    JohnG.ClearyandIanH.Witten.Datacompressionusingadaptivecodingandpartialstringmatching. IEEE Transactions on Communications, 32(4):396–402, 1984. doi: 10.1109/TCOM.1984.1096090

    arXiv 1984

  8. [8]

    Di Liberto, Claire Pelofi, Roberta Bianco, Prachi Patel, Ashesh D

    Giovanni M. Di Liberto, Claire Pelofi, Roberta Bianco, Prachi Patel, Ashesh D. Mehta, Jose L. Herrero, Alain de Cheveigné, Shihab A. Shamma, and Nima Mesgarani. Cortical encoding of melodic expectations in human temporal cortex.eLife, 9:e51784, 2020. doi: 10.7554/eLife.51784

    doi:10.7554/elife.51784 2020

  9. [9]

    Quiroga-Martinez, Niels C

    David R. Quiroga-Martinez, Niels C. Hansen, Andreas Højlund, Marcus T. Pearce, Elvira Brattico, andPeterVuust. Decomposingneuralresponsestomelodicsurpriseinmusiciansandnon-musicians: evidence for a hierarchy of predictions in the auditory system.NeuroImage, 215:116816, 2020. doi: 10.1016/j.neuroimage.2020.116816

    arXiv 2020

  10. [10]

    Corticalactivityduringnaturalistic music listening reflects short-range predictions based on long-term experience.eLife, 11:e80935,

    PiusKern,MichaHeilbron,FlorisP.deLange,andEelkeSpaak. Corticalactivityduringnaturalistic music listening reflects short-range predictions based on long-term experience.eLife, 11:e80935,

  11. [11]

    Predictive processes and the peculiar case of music

    Stefan Koelsch, Peter Vuust, and Karl Friston. Predictive processes and the peculiar case of music. Trends in Cognitive Sciences, 23(1):63–77, 2019

    2019

  12. [12]

    Zuk, Félix Bigand, Eros Quarta, Stefano Grasso, Flavia Arnese, Andrea Ravignani, Alexandra Battaglia-Mayer, and Giacomo Novembre

    Roberta Bianco, Nathaniel J. Zuk, Félix Bigand, Eros Quarta, Stefano Grasso, Flavia Arnese, Andrea Ravignani, Alexandra Battaglia-Mayer, and Giacomo Novembre. Neural encoding of musical expectations in a non-human primate.Current Biology, 34(2):444–450.e5, 2024. doi: 10.1016/j.cub.2023.12.019

    work page doi:10.1016/j.cub.2023.12.019 2024

  13. [13]

    Di Liberto, and Shihab A

    Guilhem Marion, Giovanni M. Di Liberto, and Shihab A. Shamma. The music of silence. part i: Responses to musical imagery encode melodic expectations and acoustics.The Journal of Neuroscience, 41(35):7435–7448, 2021. doi: 10.1523/JNEUROSCI.0183-21.2021

    work page doi:10.1523/jneurosci.0183-21.2021 2021

  14. [14]

    Sauvé and Marcus T

    Sarah A. Sauvé and Marcus T. Pearce. Information-theoretic modeling of perceived musical complexity.Music Perception, 37(2):165–178, 2019. doi: 10.1525/mp.2019.37.2.165

    work page doi:10.1525/mp.2019.37.2.165 2019

  15. [15]

    Gold, Marcus T

    Benjamin P. Gold, Marcus T. Pearce, Ernest Mas-Herrero, Alain Dagher, and Robert J. Zatorre. Predictability and uncertainty in the pleasure of music: a reward for learning?Journal of Neuroscience, 39(47):9397–9409, 2019

    2019

  16. [16]

    Predictive processes shape individual musical preferences.Proceedings of the National Academy of Sciences, 122(29):e2500494122, 2025

    Ernest Mas-Herrero and Josep Marco-Pallarés. Predictive processes shape individual musical preferences.Proceedings of the National Academy of Sciences, 122(29):e2500494122, 2025. doi: 10.1073/pnas.2500494122. 16

    work page doi:10.1073/pnas.2500494122 2025

  17. [17]

    Vincent K. M. Cheung, Peter M. C. Harrison, Lars Meyer, Marcus T. Pearce, John-Dylan Haynes, and Stefan Koelsch. Uncertainty and surprise jointly predict musical pleasure and amygdala, hippocampus, and auditory cortex activity.Current Biology, 29(23):4084–4092, 2019

    2019

  18. [18]

    RobertaBianco,BenjaminP.Gold,AndrewP.Johnson,andVirginiaB.Penhune.Musicpredictability and liking enhance pupil dilation and promote motor learning in non-musicians.Scientific Reports, 9(1):17060, 2019

    2019

  19. [19]

    Roberta Bianco, Peter M. C. Harrison, Mingyue Hu, Cora Bolger, Samantha Picken, Marcus T. Pearce, and Maria Chait. Long-term implicit memory for sequential auditory patterns in humans. eLife, 9:e56073, 2020

    2020

  20. [20]

    py2lispIDyOM: A Python package for the information dynamics of music (IDyOM) model.Journal of Open Source Software, 7(79):4738, 2022

    Xinyi Guan, Zeng Ren, and Claire Pelofi. py2lispIDyOM: A Python package for the information dynamics of music (IDyOM) model.Journal of Open Source Software, 7(79):4738, 2022. doi: 10.21105/joss.04738

    work page doi:10.21105/joss.04738 2022

  21. [21]

    Gold, Giovanni M

    Guilhem Marion, Fei Gao, Benjamin P. Gold, Giovanni M. Di Liberto, and Shihab A. Shamma. IDyOMpy: A new Python-based model for the statistical analysis of musical expectations.Journal of Neuroscience Methods, 415:110347, 2025. doi: 10.1016/j.jneumeth.2024.110347

    arXiv 2025

  22. [22]

    Damián H. Zanette. Zipf’s law and the creation of musical context.Musicae Scientiae, 10(1):3–18, 2006

    2006

  23. [23]

    On the emergence of zipf’s law in music.Physica A: Statistical Mechanics and its Applications, 549:124309, 2020

    Juan Ignacio Perotti and Orlando Vito Billoni. On the emergence of zipf’s law in music.Physica A: Statistical Mechanics and its Applications, 549:124309, 2020. doi: 10.1016/j.physa.2020.124309

    arXiv 2020

  24. [24]

    Heaps’ law and vocabulary richness in the history of classical music harmony.EPJ Data Science, 10(1):40, 2021

    Marc Serra-Peralta, Joan Serrà, and Álvaro Corral. Heaps’ law and vocabulary richness in the history of classical music harmony.EPJ Data Science, 10(1):40, 2021. doi: 10.1140/epjds/s13688 -021-00293-8

    work page doi:10.1140/epjds/s13688 2021

  25. [25]

    Stochastic properties of musical time series.Nature Communica- tions, 15:9280, 2024

    Corentin Nelias and Theo Geisel. Stochastic properties of musical time series.Nature Communica- tions, 15:9280, 2024. doi: 10.1038/s41467-024-53155-y

    work page doi:10.1038/s41467-024-53155-y 2024

  26. [26]

    The geometry of musical chords.Science, 313(5783):72–74, 2006

    Dmitri Tymoczko. The geometry of musical chords.Science, 313(5783):72–74, 2006

    2006

  27. [27]

    Oxford University Press, New York, 2011

    Dmitri Tymoczko.A Geometry of Music: Harmony and Counterpoint in the Extended Common Practice. Oxford University Press, New York, 2011. ISBN 9780195336672

    2011

  28. [28]

    Composingmusicwithcomplexnetworks

    XiaofanLiu,ChiK.Tse,andMichaelSmall. Composingmusicwithcomplexnetworks. InComplex Sciences–FirstInternationalConference,Complex2009,RevisedPapers,volume5ofLectureNotes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, pages 2196–2205, 2009. doi: 10.1007/978-3-642-02469-6_95

    work page doi:10.1007/978-3-642-02469-6_95 2009

  29. [29]

    On the complex network structure of musical pieces: analysis of some use cases from different music genres.Multimedia Tools and Applications, 77:16003–16029, 2018

    Stefano Ferretti. On the complex network structure of musical pieces: analysis of some use cases from different music genres.Multimedia Tools and Applications, 77:16003–16029, 2018. doi: 10.1007/s11042-017-5175-y

    work page doi:10.1007/s11042-017-5175-y 2018

  30. [30]

    Harmonic structures of Beethoven quartets: a complex network approach.The European Physical Journal B, 95(7):103, 2022

    Theo Frottier, Bertrand Georgeot, and Olivier Giraud. Harmonic structures of Beethoven quartets: a complex network approach.The European Physical Journal B, 95(7):103, 2022

    2022

  31. [31]

    Complexnetworksofharmonicstructureinclassical music

    FlorianGomez,TomLorimer,andRuediStoop. Complexnetworksofharmonicstructureinclassical music. InNonlinear Dynamics of Electronic Systems, volume 438 ofCommunications in Computer and Information Science, pages 262–269, 2014. doi: 10.1007/978-3-319-08672-9_32

    work page doi:10.1007/978-3-319-08672-9_32 2014

  32. [32]

    David, Christopher W

    Suman Kulkarni, Sophia U. David, Christopher W. Lynn, and Dani S. Bassett. Information content of note transitions in the music of J. S. Bach.Physical Review Research, 6(1):013136, 2024. 17

    2024

  33. [33]

    Ángeles Serrano

    Lluc Bono Rosselló, Robert Jankowski, Hugues Bersini, Marián Boguñá, and M. Ángeles Serrano. Trade-offs between structural richness and perceptual robustness in music network representations. Phys. Rev. Res., Jun 2026. doi: 10.1103/x6hc-jb2r. URLhttps://link.aps.org/doi/10.11 03/x6hc-jb2r

    work page doi:10.1103/x6hc-jb2r 2026

  34. [34]

    Tonalharmonyandthetopologyofdynamicalscorenetworks.Journal of Mathematics and Music, 17(2):198–212, 2023

    MarcoBuongiornoNardelli. Tonalharmonyandthetopologyofdynamicalscorenetworks.Journal of Mathematics and Music, 17(2):198–212, 2023. doi: 10.1080/17459737.2021.1969599

    arXiv 2023

  35. [35]

    Decoding the evolution of melodic and harmonic structure of western music through the lens of network science.Scientific Reports, 16(1):11121, 2026

    Niccolò Di Marco, Edoardo Loru, Alessandro Galeazzi, Matteo Cinelli, and Walter Quattrociocchi. Decoding the evolution of melodic and harmonic structure of western music through the lens of network science.Scientific Reports, 16(1):11121, 2026. doi: 10.1038/s41598-026-42872-7

    work page doi:10.1038/s41598-026-42872-7 2026

  36. [36]

    Peter M. C. Harrison, Roberta Bianco, Maria Chait, and Marcus T. Pearce. PPM-Decay: A computational model of auditory prediction with memory decay.PLoS Computational Biology, 16 (11):e1008304, 2020. doi: 10.1371/journal.pcbi.1008304

    work page doi:10.1371/journal.pcbi.1008304 2020

  37. [37]

    Methods for combining statistical models of music

    Marcus Pearce, Darrell Conklin, and Geraint Wiggins. Methods for combining statistical models of music. InInternational Symposium on Computer Music Modeling and Retrieval, pages 295–312. Springer, 2004

    2004

  38. [38]

    Hagberg, Daniel A

    Aric A. Hagberg, Daniel A. Schult, and Pieter J. Swart. Exploring network structure, dynamics, and function using networkx. In Gaël Varoquaux, Travis Vaught, and Jarrod Millman, editors, Proceedings of the 7th Python in Science Conference, pages 11–15, Pasadena, CA, 2008

    2008

  39. [39]

    Gephi: An open source software for exploring and manipulating networks

    Mathieu Bastian, Sebastien Heymann, and Mathieu Jacomy. Gephi: An open source software for exploring and manipulating networks. InProceedings of the Third International AAAI Conference on Weblogs and Social Media, pages 361–362, San Jose, CA, 2009

    2009

  40. [40]

    Pearce.The Construction and Evaluation of Statistical Models of Melodic Structure in Music Perception and Composition

    Marcus T. Pearce.The Construction and Evaluation of Statistical Models of Melodic Structure in Music Perception and Composition. PhD thesis, City University London, 2005

    2005

  41. [41]

    IRIDyOM: Exposing musical expectation as an interactive creative space

    Lluc Bono Rosselló, Julien Baudru, and Hugues Bersini. IRIDyOM: Exposing musical expectation as an interactive creative space. InProceedings of the International Conference on Computational Creativity, 2026. Forthcoming. Project repository available athttps://github.com/jbaudru/I RIDyOM

    2026

  42. [42]

    Appleton-Century-Crofts,NewYork,1971

    DanielE.Berlyne.AestheticsandPsychobiology. Appleton-Century-Crofts,NewYork,1971. ISBN 9780390086709

    1971

  43. [43]

    Routledge, 2004

    Margaret A Boden.The creative mind: Myths and mechanisms. Routledge, 2004

    2004

  44. [44]

    Geraint A. Wiggins. A preliminary framework for description, analysis and comparison of creative systems.Knowledge-Based Systems, 19(7):449–458, 2006. doi: 10.1016/j.knosys.2006.04.009

    doi:10.1016/j.knosys.2006.04.009 2006

  45. [45]

    Wiggins, and Yukie Nagai

    Tatsuya Daikoku, Geraint A. Wiggins, and Yukie Nagai. Statistical properties of musical creativity: roles of hierarchy and uncertainty in statistical learning.Frontiers in Neuroscience, 15:640412, 2021

    2021

  46. [46]

    Computational social creativity.Artificial Life, 21(3):366–378,

    Rob Saunders and Oliver Bown. Computational social creativity.Artificial Life, 21(3):366–378,

  47. [47]

    Extending the creative systems framework for the analysis of creative agent societies

    Simo Linkola and Anna Kantosalo. Extending the creative systems framework for the analysis of creative agent societies. InProceedings of the 10th International Conference on Computational Creativity, pages 204–211. Association for Computational Creativity, 2019

    2019

  48. [48]

    Emergent orchestras: A modular framework for musical robot swarms

    Lluc Bono Rosselló, Muhanad Alkilabi, Elio Tuci, Hugues Bersini, and Andreagiovanni Reina. Emergent orchestras: A modular framework for musical robot swarms. InProceedings of the 2024 Conference on Artificial Life. MIT Press, 2024. doi: 10.1162/isal_a_00746. 18

    work page doi:10.1162/isal_a_00746 2024

  49. [2015]

    doi: 10.1162/ARTL_a_00177

    work page doi:10.1162/artl_a_00177

  50. [2022]

    doi: 10.7554/eLife.80935

    doi:10.7554/elife.80935