The (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual Model
Abstract
Zhuangzi dreamt he was a butterfly and, on waking, could not tell which was the dream. We ask a version of his question of a model of the fly brain: placed inside a world generated by a world model, can a fixed brain tell that the world is a dream? We treat it as a question about what a connectome-constrained model trained on one task makes available beyond that task: do its responses make the defects of a dream linearly decodable? The observer is a published model of the fly visual system trained only for optic-flow estimation; reality is the MPI-Sintel movie on the model's 721-column retina; dreams are continuations of five real frames by three small world models (a persistence random walk, an MDN-RNN and an RSSM) at three temperatures. The judge is a classifier two-sample test on temporal statistics of cell-type responses, with sequence-level cross-validation. In the trained observer's responses all nine dreams are separable from reality (AUC 0.85-0.99, mean over 50 trained models), in those of the same network before training much less so (0.58-0.75 for the learned world models; paired difference 0.20-0.37), equally on sequences the observer never saw, and in an order unrelated to pixel error. Because these AUCs are near ceiling, we add graded probes made of each window's own real frames with only their order broken. Swapping one adjacent pair of the 14 generated frames gives 0.73 on the trained responses against 0.54 before training and 0.56 for pixel statistics; the trained curve reaches 0.75 at 1.1 swapped pairs of seven, the pixel statistics at 4.7, and the untrained network not at all. A detection horizon, the test on the first k generated frames, orders the world models where the full-length AUC saturates.
Related
- Discovering Subtypes of Neurodegenerative Progression with a Scalable Connectome-Constrained Dynamic Model
- ApproxBrain: Certified Sufficiency Frontiers in Connectome-Constrained Neural Systems
- FlyHack: Stress-Testing Mechanistic Claims in a Connectome-Constrained Visual Model
- Representational geometry as a fidelity metric for connectome-constrained networks: evidence from the Drosophila visual system
- Effect of palm oil and shea butter addition on chocolate rheology, aroma release and sensory perception: A mixture design of experiment approach
BibTeX
@misc{niimi2026butterfly,
title = {{The (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual Model}},
author = {Junichiro Niimi},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01866},
eprint = {2611.01866},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01866},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - The (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual Model
AU - Junichiro Niimi
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01866
UR - https://arxiv.org/abs/2611.01866
AB - Zhuangzi dreamt he was a butterfly and, on waking, could not tell which was the dream. We ask a version of his question of a model of the fly brain: placed inside a world generated by a world model, can a fixed brain tell that the world is a dream? We treat it as a question about what a connectome-constrained model trained on one task makes available beyond that task: do its responses make the defects of a dream linearly decodable? The observer is a published model of the fly visual system trained only for optic-flow estimation; reality is the MPI-Sintel movie on the model's 721-column retina; dreams are continuations of five real frames by three small world models (a persistence random walk, an MDN-RNN and an RSSM) at three temperatures. The judge is a classifier two-sample test on temporal statistics of cell-type responses, with sequence-level cross-validation. In the trained observer's responses all nine dreams are separable from reality (AUC 0.85-0.99, mean over 50 trained models), in those of the same network before training much less so (0.58-0.75 for the learned world models; paired difference 0.20-0.37), equally on sequences the observer never saw, and in an order unrelated to pixel error. Because these AUCs are near ceiling, we add graded probes made of each window's own real frames with only their order broken. Swapping one adjacent pair of the 14 generated frames gives 0.73 on the trained responses against 0.54 before training and 0.56 for pixel statistics; the trained curve reaches 0.75 at 1.1 swapped pairs of seven, the pixel statistics at 4.7, and the untrained network not at all. A detection horizon, the test on the first k generated frames, orders the world models where the full-length AUC saturates.
ER -