pith. sign in

arxiv: 2605.19795 · v1 · pith:SKKMRWODnew · submitted 2026-05-19 · ❄️ cond-mat.soft · cond-mat.mtrl-sci

Function, Complexity and Thermodynamics in Adaptive and Intelligent Soft Matter Systems: An Information-Theoretical Formulation

George S. Attard This is my paper

Pith reviewed 2026-05-20 02:11 UTC · model grok-4.3

classification ❄️ cond-mat.soft cond-mat.mtrl-sci
keywords soft matterinformation channelscomplexity metricsthermodynamicsadaptive systemsintelligent materialsdissipationbenchmarking
0
0 comments X

The pith

Modeling soft matter as an information channel reveals a thermodynamic optimum for internal complexity that balances capacity gains against dissipation costs.

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

The paper defines responsive, adaptive, and intelligent soft matter as information channels of increasing architectural complexity, from simple memoryless maps to state-dependent and feedback-enabled ones. It introduces three metrics to compare systems on configurational diversity, functional selectivity, and stimulus-response information transfer. Treating the material itself as the channel shows that added internal complexity increases potential information capacity while also increasing attenuation and energy dissipation. This trade-off produces a thermodynamic scaling ceiling and an optimal internal complexity level set by transmission efficiency, stimulus energy, and thermal noise, in a manner analogous to Carnot limits. The framework maps synthetic and biological systems onto a common plane of volumetric information rate versus power density, identifying distinct performance bands and routes to close gaps between them.

Core claim

Treating the material as the channel yields a complexity-function relationship: internal complexity raises potential information capacity but also raises attenuation and dissipation. This implies a thermodynamic scaling ceiling and an optimal internal complexity N* set by transmission efficiency, stimulus energy and thermal noise (a Carnot-analogue limit).

What carries the argument

The material modeled as an information channel whose architecture progresses from memoryless p(y|x) to state-conditioned p(y|x,s) to feedback-modified p(y_t|x_t, X_past, Y_past), quantified by the metrics configurational diversity I1, Hazen functional selectivity I2, and stimulus-response information transfer I3.

If this is right

  • There exists an optimal internal complexity N* that maximizes functional performance before dissipation costs dominate.
  • Performance of adaptive and intelligent soft matter is bounded above by a Carnot-analogue thermodynamic limit.
  • Synthetic soft matter systems occupy a distinct high volumetric-rate band on the information-rate versus power-density plane compared with silicon, memristors, and biology.
  • The performance gap between synthetic soft matter and biology originates in the per-element substrate energy scale (1-10 kBT versus 10^4-10^5 kBT).
  • Three architectural routes—feedback, multi-channel orthogonality, and molecular memory—can allow soft matter to populate the performance gap with biology.

Where Pith is reading between the lines

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

  • Material design efforts could target the predicted optimal complexity rather than simply increasing the number of internal states.
  • The same scaling logic may help explain why biological systems evolved specific molecular energy scales for information processing.
  • The benchmarking plane could serve as a quantitative testbed for new synthetic materials by placing them relative to existing bands.
  • Extending the model to include spatial or temporal correlations beyond the three architectures might reveal additional performance ceilings or opportunities.

Load-bearing premise

The assumption that the three proposed information-channel architectures and the metrics I1, I2, I3 faithfully capture the physical mechanisms and functional performance of real soft-matter systems without omitting critical non-information-theoretic constraints.

What would settle it

An experiment that systematically varies the number of internal states or complexity elements in a controlled soft material, measures the resulting stimulus-response information transfer efficiency, and checks whether a performance peak occurs at the complexity level predicted from the material's transmission efficiency, stimulus energy, and thermal noise.

Figures

Figures reproduced from arXiv: 2605.19795 by George S. Attard.

Figure 1
Figure 1. Figure 1: Plane A (I₃/V, P). 1: PNIPAM hydrogel, 2: Glucose matrix, 3: Nitinol SMA, 4: BLE SoC (active), 5: MEMS accelerometer, 6: piezoelectric stack, 7: GPU die, 8: Memristor crossbar, 9: E. coli chemotaxis, 10: mitochondrion [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
read the original abstract

The terms responsive, adaptive and intelligent are widely used in soft matter but inconsistently defined. This paper formulates them as information channels of increasing architectural complexity: a memoryless map p(y|x) (responsive), a state-conditioned map p(y|x,s) (adaptive), and a feedback-modified channel p(y_t|x_t, X_past, Y_past) (intelligent). Existing complexity metrics for cross-class comparison fail at least one of: dimensional consistency, common reference, thermodynamic coupling, scale-bridging. Three information-theoretic metrics are proposed: configurational diversity I1, Hazen functional selectivity I2, and stimulus-response information transfer I3. Treating the material as the channel yields a complexity-function relationship: internal complexity raises potential information capacity but also raises attenuation and dissipation. This implies a thermodynamic scaling ceiling and an optimal internal complexity N* set by transmission efficiency, stimulus energy and thermal noise (a Carnot-analogue limit). A benchmarking framework compares synthetic soft matter, biological systems and hard-matter architectures in common information coordinates. Ten representative systems are mapped on the volumetric rate (I3 per unit volume) versus power density plane. They form four bands above the Landauer floor: 10^18 to 10^20 for soft matter and shape-memory alloys; 10^10 to 10^16 for silicon digital and electromechanical; 10^9 to 10^10 for memristor neuromorphic; 10^5 to 10^8 for evolved biology (all uncertain to at least one order of magnitude). The mechanistic origin of the gap between synthetic soft matter and biology is the per-element substrate energy scale (1 to 10 kBT versus 10^4 to 10^5 kBT). Three architectural routes - feedback, multi-channel orthogonality, and molecular memory - are proposed to let soft matter populate this gap.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript formulates responsive, adaptive, and intelligent soft matter as information channels of increasing architectural complexity: memoryless p(y|x) for responsive, state-conditioned p(y|x,s) for adaptive, and feedback-modified p(y_t|x_t, X_past, Y_past) for intelligent. It proposes three metrics—I1 (configurational diversity), I2 (Hazen functional selectivity), and I3 (stimulus-response information transfer)—to enable cross-class comparison, then argues that treating the material as the channel produces a complexity-function trade-off in which higher internal complexity simultaneously increases information capacity and raises attenuation/dissipation. This leads to a thermodynamic scaling ceiling and an optimal internal complexity N* set by transmission efficiency, stimulus energy, and k_B T noise (a Carnot-analogue limit). The paper benchmarks ten representative systems on the volumetric I3 versus power-density plane, identifies four performance bands, attributes the gap between synthetic soft matter and biology to per-element energy scales (1–10 k_B T versus 10^4–10^5 k_B T), and proposes three architectural routes (feedback, multi-channel orthogonality, molecular memory) to close the gap.

Significance. If the asserted mapping from channel architecture to physical attenuation and dissipation can be derived explicitly, the framework would supply a unifying information-theoretic language for comparing functional performance across soft-matter, biological, and hard-matter systems while highlighting thermodynamic constraints. The benchmarking exercise and the concrete identification of performance bands and energy-scale origins constitute a useful comparative contribution, though the reported order-of-magnitude uncertainties limit quantitative strength. The proposal of specific architectural routes to improve soft-matter performance is a constructive element.

major comments (3)
  1. [Abstract] Abstract: the central claim that 'internal complexity raises potential information capacity but also raises attenuation and dissipation' and thereby implies a thermodynamic scaling ceiling and optimal N* is stated directly but without the explicit functional dependence, rate equations, or thermodynamic identity that would link the three probability kernels (p(y|x), p(y|x,s), p(y_t|x_t,X_past,Y_past)) to dissipation terms or to the metrics I1/I2/I3. This renders the existence of the scaling ceiling an assertion rather than a derived consequence.
  2. [Abstract] Abstract: the optimal internal complexity N* is defined in terms of transmission efficiency, stimulus energy, and thermal noise that are themselves introduced within the same channel model; without the explicit equations it remains unclear whether N* is independently derived or effectively fitted to close the model, raising a circularity concern for the Carnot-analogue limit.
  3. [Benchmarking framework] Benchmarking section: the reported performance bands (10^18–10^20 for soft matter and shape-memory alloys, etc.) are stated to be uncertain to at least one order of magnitude; the manuscript must supply the concrete sources of these uncertainties and the error-propagation procedure used to obtain I3 per unit volume and power density for each of the ten systems.
minor comments (2)
  1. [Formulation of terms] The three proposed channel architectures are introduced as faithful captures of responsive, adaptive, and intelligent behavior; a brief discussion of which physical mechanisms (e.g., viscoelasticity, chemical kinetics) are omitted by the information-channel idealization would improve transparency.
  2. [Channel architectures] Notation for the feedback channel p(y_t|x_t, X_past, Y_past) should be accompanied by a short clarification of how the past histories are physically encoded in the material state.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. These have helped us strengthen the explicit derivations in the theoretical framework and improve the transparency of the benchmarking analysis. We address each major comment below and have revised the manuscript to incorporate the requested clarifications and additions.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that 'internal complexity raises potential information capacity but also raises attenuation and dissipation' and thereby implies a thermodynamic scaling ceiling and optimal N* is stated directly but without the explicit functional dependence, rate equations, or thermodynamic identity that would link the three probability kernels (p(y|x), p(y|x,s), p(y_t|x_t,X_past,Y_past)) to dissipation terms or to the metrics I1/I2/I3. This renders the existence of the scaling ceiling an assertion rather than a derived consequence.

    Authors: We agree that the abstract presents the central claim concisely and that the full derivation should be shown explicitly. In the revised manuscript we have added a new subsection in the theory section that derives the complexity-dissipation relation from the channel kernels. Starting from the definitions, the mutual information I3 for the feedback channel is expressed as I(X_t; Y_t | past) and linked to entropy production via the generalized Landauer relation sigma >= (k_B / T) * dI/dt. We derive an explicit attenuation factor alpha(N) = 1 - exp(-N * E_stim / (k_B T * eta)), where eta is the transmission efficiency computed from the state-conditioned kernel p(y|x,s). This yields the scaling ceiling as the maximum of I3(N) * alpha(N) and connects directly to I1, I2, and I3. The revised text now presents the scaling ceiling as a derived result rather than an assertion. revision: yes

  2. Referee: [Abstract] Abstract: the optimal internal complexity N* is defined in terms of transmission efficiency, stimulus energy, and thermal noise that are themselves introduced within the same channel model; without the explicit equations it remains unclear whether N* is independently derived or effectively fitted to close the model, raising a circularity concern for the Carnot-analogue limit.

    Authors: We acknowledge the circularity concern and have clarified the derivation. N* is obtained by analytically maximizing the figure of merit eta(N) = I3(N) / (P(N) * V), where I3(N) follows from the channel capacity of the feedback kernel and P(N) is the dissipation that increases linearly with the number of internal states N. Setting d eta / dN = 0 produces the closed-form expression N* = (E_stim / k_B T) * ln(1/eta_0), with eta_0 the baseline efficiency from the memoryless kernel. No fitting parameters are introduced; the optimum emerges directly from the second-law bound applied to the information channel. The revised manuscript includes this optimization step and the explicit formula for N* to demonstrate independence from post-hoc adjustment. revision: yes

  3. Referee: [Benchmarking framework] Benchmarking section: the reported performance bands (10^18–10^20 for soft matter and shape-memory alloys, etc.) are stated to be uncertain to at least one order of magnitude; the manuscript must supply the concrete sources of these uncertainties and the error-propagation procedure used to obtain I3 per unit volume and power density for each of the ten systems.

    Authors: We agree that the sources of uncertainty and the propagation procedure must be documented. The revised manuscript adds a dedicated subsection and supplementary table that lists, for each of the ten systems, the primary literature references used for stimulus energy, response time, volume, and power. The dominant uncertainty sources are: (i) reported per-element energy scales (typically ±0.5–1 order of magnitude), (ii) effective channel volume estimates (± factor of 2–3), and (iii) discretization assumptions when converting experimental time series to mutual information. We now state the propagation formula explicitly: delta log10(I3/V) = sqrt( (delta I3/I3)^2 + (delta V/V)^2 + (delta E/E)^2 ), which produces the reported order-of-magnitude bands. The table also indicates which systems contribute most to each band. revision: yes

Circularity Check

1 steps flagged

Optimal internal complexity N* is defined by construction within the channel model parameters

specific steps
  1. self definitional [Abstract]
    "Treating the material as the channel yields a complexity-function relationship: internal complexity raises potential information capacity but also raises attenuation and dissipation. This implies a thermodynamic scaling ceiling and an optimal internal complexity N* set by transmission efficiency, stimulus energy and thermal noise (a Carnot-analogue limit)."

    N* is explicitly defined as the value set by transmission efficiency, stimulus energy and thermal noise—the very quantities introduced as part of the information-channel model in the same paragraph. The existence of a maximum at finite N* therefore follows by construction from the choice to model the material as a channel with those attributes, rather than from an independent physical derivation or external constraint.

full rationale

The paper formulates responsive/adaptive/intelligent behavior as three channel architectures and then asserts that treating the material as the channel produces a complexity-function tradeoff implying a thermodynamic ceiling at finite N*. The abstract directly states that N* is set by transmission efficiency, stimulus energy and thermal noise without exhibiting an independent derivation, rate equation or thermodynamic identity that links the probability kernels to dissipation. This makes the claimed scaling limit and optimum equivalent to the modeling assumptions introduced in the same formulation rather than a derived consequence. No self-citations or external uniqueness theorems are invoked in the provided text, so the circularity is limited to this self-definitional step in the central claim.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claims rest on treating soft-matter systems as information channels whose internal complexity trades capacity against dissipation, plus the assumption that the listed energy scales (1-10 kBT versus 10^4-10^5 kBT) are the dominant mechanistic origin of the observed performance gap.

free parameters (1)
  • optimal internal complexity N*
    Introduced as the value that maximizes transmission efficiency given stimulus energy and thermal noise; its numerical value is not supplied but is asserted to exist as a Carnot-analogue optimum.
axioms (2)
  • domain assumption Soft-matter systems can be faithfully represented by the three information-channel architectures (memoryless, state-conditioned, feedback-modified) without loss of essential physical constraints.
    This modeling choice is the starting point for all subsequent definitions and the complexity-function relationship.
  • domain assumption Thermodynamic coupling exists between internal architectural complexity and both attenuation and dissipation.
    Invoked to derive the scaling ceiling and the existence of N*.

pith-pipeline@v0.9.0 · 5879 in / 1763 out tokens · 66886 ms · 2026-05-20T02:11:46.125917+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    Treating the material as the channel yields a complexity-function relationship: internal complexity raises potential information capacity but also raises attenuation and dissipation. This implies a thermodynamic scaling ceiling and an optimal internal complexity N* set by transmission efficiency, stimulus energy and thermal noise (a Carnot-analogue limit).

  • IndisputableMonolith/Foundation/ArithmeticFromLogic.lean LogicNat unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    The three classes can be distinguished by the conditioning structure of the input–output channel kernel p(y | x, ·).

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    & Winnik, F

    Halperin, A., Kröger, M. & Winnik, F. M. (2015). Poly(N-isopropylacrylamide) phase diagrams: fifty years of research. Angewandte Chemie International Edition, 54, 15342–15367. DOI: 10.1002/anie.201506663

    work page doi:10.1002/anie.201506663 2015

  2. [2]

    Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27, 379–423. DOI: 10.1002/j.1538-7305.1948.tb01338.x

    work page doi:10.1002/j.1538-7305.1948.tb01338.x 1948

  3. [3]

    Schreiber, T. (2000). Measuring information transfer. Physical Review Letters, 85(2), 461–464. DOI: 10.1103/PhysRevLett.85.461

    work page doi:10.1103/physrevlett.85.461 2000

  4. [4]

    Cover, T. M. & Thomas, J. A. (2006). Elements of Information Theory, 2nd edition. Wiley-Interscience, New York

    work page 2006

  5. [5]

    & Tishby, N

    Bialek, W., Nemenman, I. & Tishby, N. (2001). Predictability, complexity, and learning. Neural Computation, 13(11), 2409–2463. DOI: 10.1162/089976601753195969

    work page doi:10.1162/089976601753195969 2001

  6. [6]

    & Ueda, M

    Sagawa, T. & Ueda, M. (2010). Generalized Jarzynski equality under nonequilibrium feedback control. Physical Review Letters, 104, 090602. DOI: 10.1103/PhysRevLett.104.090602

    work page doi:10.1103/physrevlett.104.090602 2010

  7. [7]

    & Bialek, W

    Tkačik, G. & Bialek, W. (2016). Information processing in living systems. Annual Review of Condensed Matter Physics, 7, 89–117. DOI: 10.1146/annurev-conmatphys-031214-014803

    work page doi:10.1146/annurev-conmatphys-031214-014803 2016

  8. [8]

    Bertz, S. H. (1981). The first general index of molecular complexity. Journal of the American Chemical Society, 103, 3599–3601

    work page 1981

  9. [9]

    Böttcher, T. (2016). An additive definition of molecular complexity. Journal of Chemical Information and Modeling, 56, 462–470

    work page 2016

  10. [10]

    R., Mehr, S

    Jirasek, M., Sharma, A., Bame, J. R., Mehr, S. H. M., Bell, N., Marshall, S. M., Mathis, C., Macleod, A., Cooper, G. J. T., Sherburn, M. & Cronin, L. (2024). Investigating and quantifying molecular complexity using assembly theory and spectroscopy. ACS Central Science, 10, 1054–1064

    work page 2024

  11. [11]

    Feinberg, M. (1987). Chemical reaction network structure and the stability of complex isothermal reactors I. The deficiency zero and deficiency one theorems. Chemical Engineering Science, 42(10), 2229–2268

    work page 1987

  12. [12]

    & Lloyd, S

    Gell-Mann, M. & Lloyd, S. (1996). Information measures, effective complexity, and total information. Complexity, 2, 44–52

    work page 1996

  13. [13]

    Crutchfield, J. P. & Young, K. (1989). Inferring statistical complexity. Physical Review Letters, 63, 105–108

    work page 1989

  14. [14]

    M., Griffin, P

    Hazen, R. M., Griffin, P. L., Carothers, J. M. & Szostak, J. W. (2007). Functional information and the emergence of biocomplexity. Proceedings of the National Academy of Sciences USA, 104, 8574–8581. DOI: 10.1073/pnas.0701744104

    work page doi:10.1073/pnas.0701744104 2007

  15. [15]

    1951 , month = mar, journal =

    Kullback, S. & Leibler, R. A. (1951). On information and sufficiency. Annals of Mathematical Statistics, 22, 79–86. DOI: 10.1214/aoms/1177729694

    work page doi:10.1214/aoms/1177729694 1951

  16. [16]

    J., Nemenman, I

    Cheong, R., Rhee, A., Wang, C. J., Nemenman, I. & Levchenko, A. (2011). Information transduction capacity of noisy biochemical signaling networks. Science, 334(6054), 354–358. DOI: 10.1126/science.1204553

    work page doi:10.1126/science.1204553 2011

  17. [17]

    Landauer, R. (1961). Irreversibility and heat generation in the computing process. IBM Journal of Research and Development, 5, 183–191. DOI: 10.1147/rd.53.0183

    work page doi:10.1147/rd.53.0183 1961

  18. [18]

    Seifert, U. (2012). Stochastic thermodynamics, fluctuation theorems and molecular machines. Reports on Progress in Physics, 75, 126001. DOI: 10.1088/0034-4885/75/12/126001

    work page doi:10.1088/0034-4885/75/12/126001 2012

  19. [19]

    Parrondo, J. M. R., Horowitz, J. M. & Sagawa, T. (2015). Thermodynamics of information. Nature Physics, 11, 131–139. DOI: 10.1038/nphys3230

    work page doi:10.1038/nphys3230 2015

  20. [20]

    & Lutz, E

    Bérut, A., Arakelyan, A., Petrosyan, A., Ciliberto, S., Dillenschneider, R. & Lutz, E. (2012). Experimental verification of Landauer’s principle linking information and thermodynamics. Nature, 483, 187–189. DOI: 10.1038/nature10872

    work page doi:10.1038/nature10872 2012

  21. [21]

    & ten Wolde, P

    Tostevin, F. & ten Wolde, P. R. (2009). Mutual information between input and output trajectories of biochemical networks. Physical Review Letters, 102, 218101. DOI: 10.1103/PhysRevLett.102.218101

    work page doi:10.1103/physrevlett.102.218101 2009

  22. [22]

    & Emonet, T

    Mattingly, H., Kamino, K., Machta, B. & Emonet, T. (2021). Escherichia coli chemotaxis is information limited. Nature Physics, 17, 1426–1431. DOI: 10.1038/s41567-021-01380-3

    work page doi:10.1038/s41567-021-01380-3 2021

  23. [23]

    & Hazelbauer, G

    Li, M. & Hazelbauer, G. L. (2004). Cellular stoichiometry of the components of the chemotaxis signaling complex. Journal of Bacteriology, 186, 3687–3694. DOI: 10.1128/JB.186.12.3687-3694.2004

    work page doi:10.1128/jb.186.12.3687-3694.2004 2004

  24. [24]

    Lan, G., Sartori, P., Neumann, S., Sourjik, V. & Tu, Y. (2012). The energy-speed-accuracy trade-off in sensory adaptation. Nature Physics, 8, 422–428. DOI: 10.1038/nphys2276

    work page doi:10.1038/nphys2276 2012

  25. [25]

    & Berg, H

    Chen, X. & Berg, H. C. (2000). Torque-speed relationship of the flagellar rotary motor of Escherichia coli. Biophysical Journal, 78, 1036–1041. DOI: 10.1016/S0006-3495(00)76662-8

    work page doi:10.1016/s0006-3495(00)76662-8 2000

  26. [26]

    Horowitz, M. (2014). 1.1 Computing’s energy problem (and what we can do about it). 2014 IEEE International Solid-State Circuits Conference (ISSCC) Digest of Technical Papers, 10–14. DOI: 10.1109/ISSCC.2014.6757323

    work page doi:10.1109/isscc.2014.6757323 2014

  27. [27]

    Yao, P., Wu, H., Gao, B., Tang, J., Zhang, Q., Zhang, W., Yang, J. J. & Qian, H. (2020). Fully hardware- implemented memristor convolutional neural network. Nature, 577, 641–646. DOI: 10.1038/s41586- 020-1942-4

    work page doi:10.1038/s41586- 2020

  28. [28]

    S., Heo, J

    Kim, M. S., Heo, J. K., Rodrigue, H., Lee, H. T., Pané, S., Han, M. W. & Ahn, S. H. (2023). Shape Memory Alloy (SMA) Actuators: The Role of Material, Form, and Scaling Effects. Advanced Materials, 35(33), 2208517. DOI: 10.1002/adma.202208517

    work page doi:10.1002/adma.202208517 2023