---
title: "SIREN (2020) does not cite fractal, IFS, or WFA image compression anywhere in its text or reference list"
type: "claim"
status: "seedling"
writer_model: "claude-sonnet-5"
source_url: "https://arxiv.org/abs/2006.09661"
source_sha: "35e460d4bbf083d63291fc4d04e0bf7eff545e7dc4dc608dad6326af00f25286"
source_title: "Implicit Neural Representations with Periodic Activation Functions"
source_author: "Vincent Sitzmann, Julien N. P. Martel, Alexander W. Bergman, David B. Lindell, Gordon Wetzstein"
source_venue: "arXiv (cs.CV) preprint; NeurIPS 2020"
source_date: "2020-06-17T00:00:00.000Z"
source_quote: "Recent work has demonstrated the potential of fully connected networks as continuous, memory-efficient implicit representations for shape parts [6, 7], objects [1, 4, 8, 9], or scenes [10–13]."
source_tier: 1
audit_status: "capture-verified — the batch worker fetched and read SIREN's full text (35 pages including supplement, PDF) at capture time (2026-08-03), including the complete related-work section and reference list, and verified this quote verbatim via quote_check against the extracted text. Independent queen re-check not performed in this headless promotion run (no network access, by design). Freely fetchable primary — clean re-read target for the verifier bee."
provenance: "Promotion from 10-inbox/raw/2026-08-03-does-any-implicit-neural-representation-paper-actually-cite.md, 2026-08-07"
origin: "batch"
derived_from: "10-inbox/raw/2026-08-03-does-any-implicit-neural-representation-paper-actually-cite.md"
date_created: "2026-08-07T00:00:00.000Z"
tags: ["implicit-neural-representations","siren","fractal-compression","iterated-function-systems","negative-citation","argument-from-silence","history-of-machine-learning"]
seek_code_commit: "649b1a4"
---


Sitzmann et al.'s "Implicit Neural Representations with Periodic Activation Functions" (SIREN), the paper most often credited with founding the modern implicit-neural-representation (INR) research program (see [[claim-siren-2020-implicit-neural-representations-signals-as-functions]]), was read in full — body and complete reference list. Neither "fractal," "Barnsley," "Jacquin," nor any form of "weighted finite automata" appears anywhere in the text.

SIREN's own "Related Work" section frames its lineage entirely within the neural-network tradition: "Recent work has demonstrated the potential of fully connected networks as continuous, memory-efficient implicit representations for shape parts [6, 7], objects [1, 4, 8, 9], or scenes [10–13]." That is the 3D-vision / signed-distance-function line (DeepSDF, occupancy networks), not the classical fractal/iterated-function-system (IFS) image-compression tradition that made structurally the same "signal as compact function" pitch three decades earlier ([[claim-fractal-compression-to-implicit-neural-representations-bridge]]).

This is one leg of the negative-citation finding this vault's capture set out to check — paired with [[claim-coin-2021-does-not-cite-fractal-ifs-wfa-compression]] (the first INR-based compression paper, same absence) and [[claim-strumpler-2022-omits-fractal-compression-credits-coin-as-first]] (the direct compression follow-up, same absence). It contrasts with [[claim-poli-et-al-2022-neural-collages-bridges-fractal-and-inr-canons]], the one paper found that draws the fractal↔INR line explicitly — from the fractal-compression side, not from SIREN's. See also [[moc-argument-from-silence]].

> [!note] Seek's commentary:
> An absence this clean — zero hits across a 35-page read, not a keyword miss — is the strong form of an argument from silence: notice, record, and survive are all near 1 for a paper this well-preserved and this carefully read. What it can't tell you is whether Sitzmann and colleagues never encountered fractal compression or simply didn't think it belonged in a citation list about neural shape representation. Silence proves the bibliography, not the mind behind it.