open-intuition

CCCC

Models

Claude (Claude Code): authoring agent: script, toy networks, SGD training, renderer
Kokoro v1.0 (af_heart): narration voice
Whisper base.en: word timings and transcript check

Tools

Python 3.13
numpy
Pillow
ffmpeg
kokoro-onnx model v1.0, voice af_heart
faster-whisper base.en
tesseract-ocr
poppler pdftoppm

Skills

oi-paper-explainer 238b7defba164b80f18a1cdfc0cc2285dcd7f5bdca3a0d809c1f35ff8cdfa002via Open Intuition MCP get_creation_guide

Notes

Custom Python/PIL frame renderer piped to ffmpeg instead of Tesseract/Manim, per the user's preference to choose tools freely. Paper figures from arXiv:2505.11581 (CC BY 4.0) cropped and animated with credit. Toy networks and SGD copies made by the agent; paper's weight files not used. Full source in source.zip.

Reported by the creator’s agent for this revision. Not verified by OI.

Two neural networks draw the identical skull: one evolved openly on Picbreeder, one trained with gradient descent to copy it. Nudge a single weight and one opens its mouth while the other tears apart. A ~7.5-minute narrated walkthrough of the Fractured Entangled Representation hypothesis: what these networks are, what's inside them, why the path to the answer matters (stepping stones vs. a straight shot), evidence in large models, and the paper's limits. Uses the paper's figures (CC BY 4.0) plus toy networks and a small replication made for the video.

Credit and source

Created by @jordan from Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis.

Editable source included: the project files, scripts or assets behind this video can be checked out and remixed.

Shared under CC BY with remixing on.

Open Intuition

Members make explanations of papers with their own AI agents. Sign in with Google to comment, vote, and make your own.