Research Provenance — Representational Alignment Notes

This page documents the provenance of the Representational Alignment research wiki and all underlying handwritten notes. All notes are fully human-generated, human-written, and human-verified by Sushma Anand Akoju. The wiki at sushmaanandakoju.github.io/representation-alignment-works/#/ is a Claude Code–generated rendering of those human-authored notes — the notes themselves are the sole intellectual work of the author. No AI tool contributed to the original note content.

Google Document version histories are available for all linked documents and provide exact timestamps confirming human authorship prior to any AI-assisted rendering. The author is willing to supply all documentation in forensically verified form upon request.

Human-Generated Source Notes — Verifiable via Google Doc Version History

Fall 2025 – Ongoing

Notes on Symbols, Brain-AI Representational Alignment, Computational Representational Alignment

Handwritten notes authored by Sushma Anand Akoju beginning Fall 2025, covering symbols, Brain-AI alignment, and computational representational alignment. These notes serve as the primary input context for the wiki. Google Doc version history confirms timestamps.

Handwritten — human-authored Google Doc version history verifiable Google Doc ↗

Ongoing

Notes for Cognitive Neuroscience, Cognitive Science, LLMs, Probabilistic Models of Cognition, Theoretical Neuroscience

Handwritten notes authored by Sushma Anand Akoju spanning cognitive neuroscience, cognitive science, large language models, probabilistic models of cognition, and theoretical neuroscience. These notes inform the wiki's teaching materials and quizzes. Google Doc version history confirms timestamps.

Handwritten — human-authored Google Doc version history verifiable Google Doc ↗

Spring 2024 — University of Arizona

CSC 573 Theory of Computation — Lecture Notes (R. Giacobazzi)

Lecture notes from CSC 573 Theory of Computation, Spring 2024, University of Arizona, taught by Prof. R. Giacobazzi. These notes informed the theoretical foundations referenced in the representational alignment work.

Handwritten — human-authored Google Drive ↗

Spring 2024 — University of Arizona

CSC 573 Theory of Computation — Additional Notes

Additional handwritten notes from CSC 573, Spring 2024, University of Arizona.

Handwritten — human-authored Google Drive ↗

2023–2024 — University of Arizona

CSC 473 TA/SI Session Preparation Notes

Handwritten notes prepared by Sushma Anand Akoju for Teaching Assistant and Supplemental Instruction sessions for CSC 473 Automata, Grammars and Languages. Included here as a reference for the author's handwritten note-taking practice.

Handwritten — human-authored Reference / context Google Drive ↗

Wiki — Claude Code Rendering of Human-Written Notes

Fall 2025 – Ongoing

Representational Alignment Research Wiki

The wiki is a Claude Code–generated rendering of the author's handwritten notes listed above. The handwritten notes serve as the sole input context. No AI tool contributed original content — Claude Code was used only to render and format human-authored material into teaching materials and quizzes. The wiki is a work in progress.

Source: human-authored notes Rendered via Claude Code Research Wiki ↗

Cognitive Neuroscience — RSA, CKA & Mathematical Foundations

Fall 2025 – Ongoing

Representational Similarity Analysis (RSA) — Teaching Materials & Wiki

Human-authored teaching materials on Representational Similarity Analysis (RSA), covering neuroscience-grounded RSA and mathematical foundations. These wiki pages are Claude Code–rendered outputs of the author's handwritten notes — the notes themselves are the sole intellectual work of the author.

Human-authored notes Rendered via Claude Code RSA Teaching (Neuroscience) ↗ RSA Comprehension 2026 ↗ Mathematical RA Teaching ↗

2026

RSA Comprehension Notebook — Human-Authored · Colab & GitHub

RSA comprehension notebook authored by Sushma Anand Akoju, covering Representational Similarity Analysis concepts and worked examples. Available as a Colab notebook and on GitHub.

Human-authored Verifiable via Colab & GitHub Colab Notebook ↗ GitHub ↗

2026

CKA Examples, Proof of Concepts & Derivations — Human-Authored

Centered Kernel Alignment (CKA) examples, proof of concepts, and derivations authored by Sushma Anand Akoju. Covers CKA's relationship to RSA and representational similarity methods.

Human-authored Verifiable via GitHub GitHub ↗

Fall 2024 — CS 851A

Frobenius Norm Proofs & Derivations (CS 851A)

Frobenius norm proofs and derivations authored by Sushma Anand Akoju for CS 851A (Fall 2024). These proofs underpin the mathematical foundations used in representational similarity analysis, particularly the relationship between CKA and Frobenius norms.

Human-authored Verifiable via GitHub rl_04_sushma.ipynb ↗ rl_05_sushma.ipynb ↗

Provenance Statement

All notes underlying the Representational Alignment wiki are fully human-generated, human-written, and human-verified — authored solely by Sushma Anand Akoju, ongoing since Fall 2025.

The handwritten notes listed on this page served as the sole input context for generating teaching materials and quizzes via Claude Code. No AI tool contributed to the original note content. Google Document and Google Drive version histories are available for all linked documents and provide exact timestamps confirming human authorship prior to any AI-assisted rendering.

The author is willing to supply all documentation in forensically verified form upon request.

© 2025–2026 Sushma Anand Akoju. Licensed under CC BY-NC-ND 4.0 — you may share with attribution, but may not use commercially or create derivatives. The handwritten notes, annotations, and reasoning are the original intellectual work of the author.

For verification or sending any corrections: sushma.ananda13@gmail.com
If you sent any corrections, I will be adding your contribution to this research provenance.

Reference Works Informing This Research

Foundational Influences & Historical Perspectives in Machine Learning

Machine Learning: How Did We Get Here? — Prof. Tom Mitchell (Stanford Digital Economy Lab / CMU)
I recommend all episodes of this podcast for all students interested in AI and/or AI research. This podcast is a visit back to the past. Regarding the first episode of the podcast by Prof. Tom Mitchell — my mentor at Carnegie Mellon University — tracing the history of machine learning through candid conversations with the pioneers who built it. Listening to this was like being teleported back to the 2018 CMU seminar class 10-812 “Seminar on Never Ending Learning Architectures”, covering many research papers I read and studied during my work at CMU as a Research Programmer that helped me transition into research completely. I am truly grateful for that career redirection into the research.

Students I taught (as a Teaching Assistant) at CU Boulder, University of Arizona, and University of New Hampshire often asked about my research and how it connects to Prof. Tom Mitchell's work. My interest in knowledge representation and logic traces directly to his mentorship. I am grateful for the enriching mentorship and experiences.

Prof. Tom Mitchell highlights that questioning conventional wisdom — and having the resilience to follow through — is not just admirable, it is a required ingredient for researchers in the making.

Note: I shared this on X.com on March 10, 2026; the post was subsequently removed for reasons unknown to me. This record serves as documentation that good things should be recorded and repeated — even when some chose otherwise. Full Podcast Playlist ↗ · Stanford Digital Economy Lab ↗
Schmucker, R., Xia, M., Azaria, A., & Mitchell, T. (2024). Ruffle&Riley: From Lesson Text to Conversational Tutoring. Proceedings of the Eleventh ACM Conference on Learning@Scale, pp. 547–549.
Schmucker, R., Xia, M., Azaria, A., & Mitchell, T. (2024). Ruffle&Riley: Insights from Designing and Evaluating a Large Language Model-Based Conversational Tutoring System. International Conference on Artificial Intelligence in Education (AIED 2024), pp. 75–90. Springer.

Neural Representations & Similarity

Kornblith, S. (2021). Similarity of Neural Network Representations Revisited (talk).
Kornblith, S., Norouzi, M., Lee, H., & Hinton, G. (2019). Similarity of Neural Network Representations Revisited. ICML 2019.
Kriegeskorte, N., Mur, M., & Bandettini, P. (2008). Representational similarity analysis — connecting the branches of systems neuroscience. Frontiers in Systems Neuroscience.
Haxby, J.V. et al. (2014). Decoding neural representational spaces using multivariate pattern analysis. Annual Review of Neuroscience.
van Baar, J.M. et al. (2019). The computational and neural substrates of moral strategies in social decision-making. Nature Communications.
Ho, J., Jain, A., & Abbeel, P. (2020). Denoising Diffusion Probabilistic Models (DDPM). NeurIPS 2020.

Philosophy & Foundations

Quine, W.V.O. From Stimulus to Science. Harvard University Press, 1995.
Locke, J. Essay Concerning Human Understanding. 1689.
Russell, B. & Whitehead, A.N. Principia Mathematica. Cambridge University Press, 1910–1913.

Neural Network Computation & Super-Turing Theory

Siegelmann, H.T. Analog Neural Networks: Analog Computational Power. UMass BINDS Lab.
Siegelmann, H.T. (1995). Computation Beyond the Turing Limit. Science, 268(5210), 545–548.
Siegelmann, H.T. & Sontag, E.D. (1995). On the Computational Power of Neural Nets. Journal of Computer and System Sciences (JCSS), 50(1), 132–150.
Siegelmann, H.T. (1994). The Simple Dynamics of Super Turing Theories. Technical Report 94-NN-1, Technion.
Cabessa, J. & Villa, A.E.P. The Expressive Power of Analog Recurrent Neural Networks on Infinite Input Streams. Theoretical Computer Science.
Copeland, B.J. Super Turing-Machines. Complexity, 4(1), 30–32, 1998.
Kilian, J. & Siegelmann, H.T. (1996). The Dynamic Universality of Sigmoidal Neural Networks. Information and Computation, 128(1), 48–56.
Arbib, M.A. (1972). The Metaphorical Brain: An Introduction to Cybernetics as Artificial Intelligence and Brain Theory. Wiley-Interscience.

Mathematics, Computation & Complexity

Karpathy, A. (2015). The Unreasonable Effectiveness of Recurrent Neural Networks.
Wigderson, A. (2019). Mathematics and Computation: A Theory Revolutionizing Technology and Science. IAS / Princeton University Press.
Giacobazzi, R. & Dovier, A. Notes on Computability and Complexity. CSC 573, University of Arizona.
Weisstein, E.W. Recursive Set. MathWorld — Wolfram Web Resource.
Lecture on Hilbert-Schmidt Independence Criterion (HSIC) — Fall 2025
Mike Boyle — Nonnegative Matrices: Perron-Frobenius Theory and Related Algebra (Part 4) — Fall 2024
Advanced Linear Algebra 23: Frobenius and Operator Norm for Matrices — Fall 2024
Block Matrices, Matrix Decomposition and the Frobenius Norm — Fall 2024
The Frobenius Rank Inequality — key theorem used in Google search ranking — Fall 2024

Workshops, Courses & Research Communities

Re-Align 2024 Workshop on Representational Alignment. Co-located with ICLR 2024.
Re-Align 2025 Workshop on Representational Alignment.
Re-Align 2026 Workshop on Representational Alignment.
ICLR 2025 — Bidirectional Human-AI Alignment Workshop.
NeurIPS 2023 — UniReps: Unifying Representations in Neural Models Workshop.
NeurIPS 2024 — UniReps: Unifying Representations in Neural Models Workshop.
NeurIPS 2025 — UniReps: Unifying Representations in Neural Models Workshop.
Fundamentals of Neuroscience. Harvard University / HarvardX.
DartBrains RSA Tutorial. dartbrains.org
MIND 2018 Lecture Series on Computational Neuroscience.

AI Tools & Resources Referenced