| title | Prediction Bottlenecks Don't Discover Causal Structure (But Here's What They Actually Do) |
|---|---|
| authors | Ankit Hemant Lade, Sai Krishna Jasti, Indar Kumar, Aman Chadha |
| date | 2026-05-09 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.09169 |
| links | abs pdf hf |
| abstract | A Mamba state-space model trained only for next-step prediction appears to recover Granger-causal structure through a simple readout S = |W_{out} W_{in}|, with early experiments suggesting the phenomenon generalized across architectures and benefited from interventional data at p < 10^{-5}. We package the protocol used to test that claim -- standardized synthetic generators (VAR/Lorenz/CauseMe-style), three intervention semantics (do(X=c), soft-noise, random-forcing), edge-provenance cards on three real datasets, and size-matched control arms -- as a reusable falsification benchmark, and walk the claim through it in five stages. The method-level claim does not survive: (i) a plain linear bottleneck does as well or better; (ii) tuned Lasso beats the bottleneck on synthetic CauseMe-style benchmarks, and on Lorenz-96 (the only real benchmark with unambiguous ground truth) classical PCMCI and Granger lead a tight cluster in which the bottleneck trails; (iii) the headline intervention advantage is roughly 60% a sample-size confound, and the residual disappears under standard do(X=c) interventions, surviving only under a non-standard random-forcing scheme; (iv) even that residual reproduces, with a larger effect, in classical bivariate Granger -- the effect is method-agnostic. What survives is a narrow characterization result; the benchmark is the lasting artifact, and each stage above is one of its control arms. |
cite
@misc{lade2026prediction,
title={Prediction Bottlenecks Don't Discover Causal Structure (But Here's What They Actually Do)},
author={Ankit Hemant Lade and Sai Krishna Jasti and Indar Kumar and Aman Chadha},
year={2026},
eprint={2605.09169},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.09169}
}