Skip to content
Preprint · CC-BY via OpenAlex

Self-Supervised Behavioral Representations Across the Life Course: A Killifish Case Study

Jen-Chien Chang, Teruhisa Komatsu, Shuichi Onami

bioRxiv (Cold Spring Harbor Laboratory) · 2026

Abstract

Abstract Self-supervised foundation models of aging are increasingly built from longitudinal data (biobanks, electronic health records, wearables) that is inherently incomplete: no individual is followed across a whole lifetime, and how much of each life is captured varies widely. This raises two linked questions: is it worth modeling an individual’s whole life course rather than its current state, and can such a model be built from brief, fragmentary records? No human cohort can settle them, because none offers a complete life to compare against. We turn to the African turquoise killifish ( Nothobranchius furzeri ), tracked from youth to natural death in publicly released recordings, as a controlled testbed: its complete lifespans provide the full-life reference that human data lacks. On these data we build LifeMAE, a two-stage self-supervised model: a day encoder that summarizes each day of behavior, then a life-course encoder over the trajectory of those daily summaries. We find that the day encoder alone is already strong: from a single day of behavior it predicts chronological age, separates long-from short-lived individuals (coarsely), and flags nearness to death. Adding the life-course encoder improves on none of the three; each is matched by trivially aggregating the day-level predictions (a smoother for age, an early-life average for lifespan). Near-term mortality seems the exception, where the whole-life model looks far better (AUROC 0.81 to 0.91), but the gain is not behavioral: it reflects where each day falls within the observation window (a cue supplied by the model’s encoding of time), and a single-day model given that cue closes the gap at any observation length. For these traits, an individual’s place in its life course is legible from a single day: the trajectory stage is unnecessary, and the record it needs is as short as one day, the finest grain our day-level setup resolves. For characterizing a cohort, this favors observing many individuals briefly over tracking a few for long. The result joins a growing body of work in which deep and foundation models, fairly benchmarked, fail to beat deliberately simple baselines. We add a concrete mechanism for the over-optimism: a model’s encoding of time can leak the very quantity it predicts, which backward-looking evaluation mistakes for learned biology, so only evaluation fixed to the moment of prediction is trustworthy.

◌ CITATION ONLY
Full text is not openly licensed for redistribution here. Read it at the source:

Read at source →

Provenance

Source
OpenAlex
DOI
10.64898/2026.06.23.733896
Canonical
link ↗
Fetched
2026-11-15 MST

Cite this

APA
Chang, J., Komatsu, T., &amp; Onami, S. (2026). Self-Supervised Behavioral Representations Across the Life Course: A Killifish Case Study. <em>bioRxiv (Cold Spring Harbor Laboratory)</em>. https://doi.org/10.64898/2026.06.23.733896
Vancouver
Chang J, Komatsu T, Onami S. Self-Supervised Behavioral Representations Across the Life Course: A Killifish Case Study. bioRxiv (Cold Spring Harbor Laboratory). 2026. doi:10.64898/2026.06.23.733896.
BibTeX
@unpublished{jenchien2026SelfSu, title = {Self-Supervised Behavioral Representations Across the Life Course: A Killifish Case Study}, author = {Jen-Chien Chang and Teruhisa Komatsu and Shuichi Onami}, journal = {bioRxiv (Cold Spring Harbor Laboratory)}, year = {2026}, doi = {10.64898/2026.06.23.733896}, }

Research neighborhood

References, citing works, and semantically nearest findings. Click a node to open it.

Related findings