alex/copycat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
b5323b5a74d0e4d2eb4d6884fec23d56ef6a2c30
copycat/LaTeX/paper2/paper.aux
Alex Linhares b5323b5a74 Add paper 2: information-theoretic unification of workspace graph metrics 
Introduces a 21-page paper focused exclusively on the Copycat Workspace,
proposing three graph-metric substitutions for hardcoded constants and
unifying them under a common information-theoretic framework:

- Clustering coefficient as support vector (replaces 0.6^(1/n^3) factor)
- Betweenness centrality as salience weights (replaces fixed 0.2/0.8 ratios)
- Subgraph density as length factors (replaces step function {5,20,60,90})

Central contribution: all three metrics are ratios actual/maximum in [0,1],
interpretable as probabilities, yielding Shannon self-information in bits.
This common unit enables principled arithmetic via workspace entropy
H(W) = sum(I_C) + sum(I_B) + sum(I_rho), with the conjecture T ∝ H(W).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-27 17:56:54 +00:00

165 lines
17 KiB
TeX
Raw Blame History

\relax
\providecommand\hyper@newdestlabel[2]{}
\providecommand\HyField@AuxAddToFields[1]{}
\providecommand\HyField@AuxAddToCoFields[2]{}
\citation{gentner1983structure,hofstadter1995fluid}
\citation{mitchell1993analogy}
\@writefile{toc}{\contentsline {section}{\numberline {1}Introduction}{1}{section.1}\protected@file@percent }
\citation{linhares2025graph}
\@writefile{toc}{\contentsline {subsection}{\numberline {1.1}Organization of This Paper}{2}{subsection.1.1}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {2}The Copycat Workspace: Background and Hardcoded Mechanisms}{2}{section.2}\protected@file@percent }
\newlabel{sec:workspace}{{2}{2}{The Copycat Workspace: Background and Hardcoded Mechanisms}{section.2}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {2.1}The Workspace as a Dynamic Graph}{2}{subsection.2.1}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Nodes.}{3}{section*.1}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Edges.}{3}{section*.2}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Dynamics.}{3}{section*.3}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {2.2}Three Hardcoded Mechanisms and Their Limitations}{3}{subsection.2.2}\protected@file@percent }
\@writefile{lot}{\contentsline {table}{\numberline {1}{\ignorespaces The three hardcoded mechanisms examined in this paper, with their locations, current formulations, and proposed graph-metric replacements.}}{3}{table.1}\protected@file@percent }
\newlabel{tab:hardcoded}{{1}{3}{The three hardcoded mechanisms examined in this paper, with their locations, current formulations, and proposed graph-metric replacements}{table.1}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {2.3}Why Adaptivity Matters}{3}{subsection.2.3}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {3}Clustering Coefficient as Support Vector}{4}{section.3}\protected@file@percent }
\newlabel{sec:clustering}{{3}{4}{Clustering Coefficient as Support Vector}{section.3}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.1}Current Bond External Strength}{4}{subsection.3.1}\protected@file@percent }
\newlabel{eq:supportfactor}{{1}{4}{Current Bond External Strength}{equation.1}{}}
\@writefile{lol}{\contentsline {lstlisting}{\numberline {1}Bond external strength in bond.py}{4}{lstlisting.1}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {3.2}The Local Clustering Coefficient}{4}{subsection.3.2}\protected@file@percent }
\newlabel{eq:clustering}{{2}{4}{Local Clustering Coefficient}{equation.2}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.3}Defining the Support Vector}{4}{subsection.3.3}\protected@file@percent }
\newlabel{eq:supportvector}{{3}{4}{Support Vector}{equation.3}{}}
\newlabel{eq:bondclustering}{{4}{5}{Defining the Support Vector}{equation.4}{}}
\newlabel{eq:bondexternal}{{5}{5}{Defining the Support Vector}{equation.5}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {3.4}Properties of the Support Vector}{5}{subsection.3.4}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Boundedness.}{5}{section*.4}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Monotone local coherence.}{5}{section*.5}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Self-normalization.}{5}{section*.6}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Elimination of arbitrary constants.}{5}{section*.7}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Computational complexity.}{5}{section*.8}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {3.5}The Support Vector as a Workspace Fingerprint}{5}{subsection.3.5}\protected@file@percent }
\@writefile{loa}{\contentsline {algorithm}{\numberline {1}{\ignorespaces Compute Support Vector}}{6}{algorithm.1}\protected@file@percent }
\newlabel{alg:supportvector}{{1}{6}{The Support Vector as a Workspace Fingerprint}{algorithm.1}{}}
\@writefile{toc}{\contentsline {section}{\numberline {4}Betweenness Centrality as Salience Weights}{6}{section.4}\protected@file@percent }
\newlabel{sec:betweenness}{{4}{6}{Betweenness Centrality as Salience Weights}{section.4}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.1}Current Salience Computation}{6}{subsection.4.1}\protected@file@percent }
\newlabel{eq:intrasalience}{{6}{6}{Current Salience Computation}{equation.6}{}}
\newlabel{eq:intersalience}{{7}{6}{Current Salience Computation}{equation.7}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.2}Betweenness Centrality}{6}{subsection.4.2}\protected@file@percent }
\newlabel{eq:betweenness}{{8}{6}{Betweenness Centrality}{equation.8}{}}
\newlabel{eq:normbetweenness}{{9}{7}{Normalized Betweenness Centrality}{equation.9}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.3}Betweenness as Salience Weight}{7}{subsection.4.3}\protected@file@percent }
\newlabel{eq:intrabs}{{10}{7}{Betweenness as Salience Weight}{equation.10}{}}
\newlabel{eq:interbs}{{11}{7}{Betweenness as Salience Weight}{equation.11}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.4}Cognitive and Structural Motivation}{7}{subsection.4.4}\protected@file@percent }
\citation{brandes2001faster}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.5}Formal Properties of Betweenness Salience}{8}{subsection.4.5}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {4.6}Computational Considerations}{8}{subsection.4.6}\protected@file@percent }
\newlabel{eq:saliencevector}{{12}{8}{Computational Considerations}{equation.12}{}}
\@writefile{loa}{\contentsline {algorithm}{\numberline {2}{\ignorespaces Compute Betweenness Salience Weights}}{8}{algorithm.2}\protected@file@percent }
\newlabel{alg:betweenness}{{2}{8}{Computational Considerations}{algorithm.2}{}}
\@writefile{toc}{\contentsline {section}{\numberline {5}Subgraph Density as Length Factor}{9}{section.5}\protected@file@percent }
\newlabel{sec:density}{{5}{9}{Subgraph Density as Length Factor}{section.5}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {5.1}Current Length Factor Step Function}{9}{subsection.5.1}\protected@file@percent }
\newlabel{eq:stepfunction}{{13}{9}{Current Length Factor Step Function}{equation.13}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {5.2}Induced Subgraph Density}{9}{subsection.5.2}\protected@file@percent }
\newlabel{eq:density}{{14}{9}{Induced Subgraph Density}{equation.14}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {5.3}Subgraph Density as a Continuous Length Factor}{9}{subsection.5.3}\protected@file@percent }
\newlabel{eq:densitylf}{{15}{9}{Subgraph Density as a Continuous Length Factor}{equation.15}{}}
\newlabel{eq:chaindensity}{{16}{9}{}{equation.16}{}}
\newlabel{eq:sizedensity}{{17}{10}{Subgraph Density as a Continuous Length Factor}{equation.17}{}}
\@writefile{lot}{\contentsline {table}{\numberline {2}{\ignorespaces Current step function versus density-based length factor for chain groups (path graphs $P_n$). The size-weighted variant $\rho \times n$ is constant for all chain groups, reflecting that all chains provide the same ratio of edges to nodes.}}{10}{table.2}\protected@file@percent }
\newlabel{tab:density}{{2}{10}{Current step function versus density-based length factor for chain groups (path graphs $P_n$). The size-weighted variant $\rho \times n$ is constant for all chain groups, reflecting that all chains provide the same ratio of edges to nodes}{table.2}{}}
\@writefile{loa}{\contentsline {algorithm}{\numberline {3}{\ignorespaces Compute Subgraph Density Length Factor}}{10}{algorithm.3}\protected@file@percent }
\newlabel{alg:density}{{3}{10}{Subgraph Density as a Continuous Length Factor}{algorithm.3}{}}
\@writefile{toc}{\contentsline {section}{\numberline {6}Worked Example: Tracing Metrics Through an Analogy Problem}{10}{section.6}\protected@file@percent }
\newlabel{sec:example}{{6}{10}{Worked Example: Tracing Metrics Through an Analogy Problem}{section.6}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {6.1}Early Workspace: No Bonds Built}{11}{subsection.6.1}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {6.2}Mid Workspace: Bonds within Target String}{11}{subsection.6.2}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {6.3}Late Workspace: Groups and Correspondences Formed}{11}{subsection.6.3}\protected@file@percent }
\@writefile{lof}{\contentsline {figure}{\numberline {1}{\ignorespaces Evolution of the three workspace metric vectors during a representative run on the problem \texttt {abc}$\to $\texttt {abd}, \texttt {ppqqrr}$\to $\texttt {?}. Top panel: support vector components $s_i = C(v_i)$; middle panel: salience weights $\hat {B}(v_i)$; bottom panel: group density $\rho (G_S)$ for all recognized groups. Vertical dashed lines mark bond proposal (B), group formation (G), and correspondence (C) events. As structure accumulates, the salience vector concentrates on the central group while the density vector saturates for individual groups.}}{12}{figure.1}\protected@file@percent }
\newlabel{fig:workspace_metrics}{{1}{12}{Evolution of the three workspace metric vectors during a representative run on the problem \texttt {abc}$\to $\texttt {abd}, \texttt {ppqqrr}$\to $\texttt {?}. Top panel: support vector components $s_i = C(v_i)$; middle panel: salience weights $\hat {B}(v_i)$; bottom panel: group density $\rho (G_S)$ for all recognized groups. Vertical dashed lines mark bond proposal (B), group formation (G), and correspondence (C) events. As structure accumulates, the salience vector concentrates on the central group while the density vector saturates for individual groups}{figure.1}{}}
\@writefile{lot}{\contentsline {table}{\numberline {3}{\ignorespaces Summary of three metrics at three workspace snapshots for the \texttt {ppqqrr} problem. Mean support ($\bar {s}$), maximum betweenness ($\max _v \hat {B}(v)$), and mean group density ($\bar {\rho }$) track workspace evolution.}}{12}{table.3}\protected@file@percent }
\newlabel{tab:example}{{3}{12}{Summary of three metrics at three workspace snapshots for the \texttt {ppqqrr} problem. Mean support ($\bar {s}$), maximum betweenness ($\max _v \hat {B}(v)$), and mean group density ($\bar {\rho }$) track workspace evolution}{table.3}{}}
\@writefile{toc}{\contentsline {section}{\numberline {7}From Ratios to Probabilities: An Information-Theoretic Unification}{12}{section.7}\protected@file@percent }
\newlabel{sec:information}{{7}{12}{From Ratios to Probabilities: An Information-Theoretic Unification}{section.7}{}}
\citation{erdos1959random}
\citation{boccaletti2006complex}
\citation{shannon1948mathematical,cover2006elements}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.1}All Three Metrics as Ratios to Maximum}{13}{subsection.7.1}\protected@file@percent }
\newlabel{eq:cc_ratio}{{18}{13}{All Three Metrics as Ratios to Maximum}{equation.18}{}}
\newlabel{eq:bc_ratio}{{19}{13}{All Three Metrics as Ratios to Maximum}{equation.19}{}}
\newlabel{eq:rho_ratio}{{20}{13}{All Three Metrics as Ratios to Maximum}{equation.20}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.2}The Probabilistic Interpretation}{13}{subsection.7.2}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {7.3}Shannon Self-Information}{13}{subsection.7.3}\protected@file@percent }
\newlabel{eq:shannon}{{21}{13}{Shannon Self-Information}{equation.21}{}}
\newlabel{eq:ic}{{22}{13}{Shannon Self-Information}{equation.22}{}}
\newlabel{eq:ib}{{23}{13}{Shannon Self-Information}{equation.23}{}}
\newlabel{eq:irho}{{24}{13}{Shannon Self-Information}{equation.24}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.4}Interpreting the Information Content}{14}{subsection.7.4}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Support information $I_C(b)$.}{14}{section*.9}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Salience information $I_B(v)$.}{14}{section*.10}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Length information $I_\rho (S)$.}{14}{section*.11}\protected@file@percent }
\newlabel{eq:chaininfo}{{25}{14}{Length information $I_\rho (S)$}{equation.25}{}}
\@writefile{lot}{\contentsline {table}{\numberline {4}{\ignorespaces Information content $I = -\log _2 p$ (bits) for representative probability values. The same table applies to all three metrics since they share the same domain $[0,1]$.}}{14}{table.4}\protected@file@percent }
\newlabel{tab:infobits}{{4}{14}{Information content $I = -\log _2 p$ (bits) for representative probability values. The same table applies to all three metrics since they share the same domain $[0,1]$}{table.4}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.5}Bits as a Common Dimension}{14}{subsection.7.5}\protected@file@percent }
\newlabel{eq:workspace_entropy}{{26}{15}{Bits as a Common Dimension}{equation.26}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.6}Workspace Entropy and Temperature}{15}{subsection.7.6}\protected@file@percent }
\newlabel{eq:tempentropy}{{27}{15}{Speculative}{equation.27}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {7.7}The Additive Property and Its Consequences}{15}{subsection.7.7}\protected@file@percent }
\citation{jaynes1957information}
\citation{bianconi2009entropy}
\@writefile{toc}{\contentsline {paragraph}{Budgeting.}{16}{section*.12}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Comparison.}{16}{section*.13}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Temperature.}{16}{section*.14}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {7.8}Maximum Entropy Interpretation}{16}{subsection.7.8}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {8}Discussion}{16}{section.8}\protected@file@percent }
\newlabel{sec:discussion}{{8}{16}{Discussion}{section.8}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {8.1}Theoretical Implications}{16}{subsection.8.1}\protected@file@percent }
\citation{shannon1948mathematical}
\citation{cover2006elements}
\@writefile{toc}{\contentsline {subsection}{\numberline {8.2}Empirical Predictions}{17}{subsection.8.2}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Prediction 1: Workspace Entropy Correlates with Temperature.}{17}{section*.15}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Prediction 2: Information-Based Urgency Outperforms Fixed Weights.}{17}{section*.16}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Prediction 3: Entropy Decreases Monotonically in Successful Runs.}{17}{section*.17}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Prediction 4: High $I_C$ Bonds Are Broken More Frequently.}{17}{section*.18}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Prediction 5: Final Workspace Entropy Is Invariant to Problem Difficulty.}{17}{section*.19}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {8.3}Connections to Thermodynamics}{17}{subsection.8.3}\protected@file@percent }
\citation{rashevsky1955life}
\citation{mowshowitz1968entropy}
\citation{dehmer2011history}
\citation{newman2018networks}
\citation{jaynes1957information}
\citation{watts1998collective}
\newlabel{eq:boltzmann}{{28}{18}{Connections to Thermodynamics}{equation.28}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {8.4}Connections to Related Work}{18}{subsection.8.4}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Graph entropy.}{18}{section*.20}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Betweenness in cognitive systems.}{18}{section*.21}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Information theory and cognition.}{18}{section*.22}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Small-world topology and clustering.}{18}{section*.23}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {8.5}Limitations and Open Questions}{19}{subsection.8.5}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Zero probabilities.}{19}{section*.24}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Independence assumption.}{19}{section*.25}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Graph size dependence.}{19}{section*.26}\protected@file@percent }
\@writefile{toc}{\contentsline {paragraph}{Direction of causality.}{19}{section*.27}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {9}Conclusion}{19}{section.9}\protected@file@percent }
\newlabel{sec:conclusion}{{9}{19}{Conclusion}{section.9}{}}
\citation{linhares2025graph}
\bibstyle{plain}
\bibdata{references}
\bibcite{bianconi2009entropy}{1}
\bibcite{boccaletti2006complex}{2}
\bibcite{brandes2001faster}{3}
\bibcite{cover2006elements}{4}
\bibcite{dehmer2011history}{5}
\bibcite{erdos1959random}{6}
\bibcite{gentner1983structure}{7}
\bibcite{hofstadter1995fluid}{8}
\bibcite{jaynes1957information}{9}
\bibcite{linhares2025graph}{10}
\bibcite{mitchell1993analogy}{11}
\bibcite{mowshowitz1968entropy}{12}
\bibcite{newman2018networks}{13}
\bibcite{rashevsky1955life}{14}
\bibcite{shannon1948mathematical}{15}
\bibcite{watts1998collective}{16}
\gdef \@abspage@last{21}
Reference in New Issue View Git Blame Copy Permalink