Immediate Derivatives Suffice for Online Recurrent Adaptation

arXiv:2603.28750v3 Announce Type: replace Abstract: For three decades online recurrent learning has been assumed to require propagating a Jacobian tensor through the network's dynamics at O(n^4) per step. We show it doesn't. Dropping the propagation entirely (d=0, O(n^2) memory) matches full RTRL within CI on held-out BCI cross-session drift (TOST equivalent within pm 3 pp at n=20, Adam, float64), and across vanilla-RNN synthetic cells (sine and Lorenz under Adam and SGD) and LSTM/sine under Adam. A decomposition g_{RTRL} = g_{imm} + g_{past} explains why. On BCI, g_{past} concentrates in a single direction (top-1 singular fraction 0.62-0.74 across four optimizers, vs 0.333 for g_{imm}), and the four-optimizer full-RTRL-vs-d=0 recovery gap tracks each optimizer's per-layer update-magnitude ratio |Delta W_{hh}|/|Delta W_{out}| monotonically. A stationary (no-drift) control collapses both concentrations to ~0.6: the drift-specific signal is the differential, not g_{past}'s absolute rank-1 structure. The signature and the behavioral gap both collapse on LSTM, consistent with a mechanism specific to additive linear recurrence. On synthetic sine, g_{imm} is redundant with g_{past}, which predicts the synthetic null. Full RTRL's one robust advantage is LARS (+17 to +27 pp), but d=0+LARS also fails to adapt independently; the gap is an optimizerimesmethod interaction, not a method-quality claim. We characterize the regime: d=0+Adam+float64 is robust; SGD, Adafactor, and float32 have specific fragilities documented in the paper. On the evaluated cells, the 1000imes memory saving at n=1024 (O(n^2) vs O(n^4)) comes with no measured recovery cost.

Source: arXiv cs.LG | 2026-04-29

Related