ICML 2026 · Conference paper
Global Credit Assignment via Dynamical Criticality
Criticality-driven Online Local Alignment exploits long-range critical correlations so local recurrent updates approximate global backpropagation in constant activation memory.
The scientific question
Can dynamical criticality allow strictly local learning rules to achieve global credit assignment?
Backpropagation through time requires memory proportional to sequence length, whereas standard biological local rules suffer from severe gradient approximation bias.

The core idea
Credit Propagation via Criticality
COLA places recurrent networks near dynamical criticality, using intrinsic long-range spatiotemporal correlations to propagate error information across time through a strictly local online update rule.
What we found
Sequence-Independent Memory and Stability
With only O(H) auxiliary state and constant activation memory independent of sequence length, COLA is competitive with backpropagation through time on standard benchmarks and demonstrates superior robustness on stability-sensitive tasks.
Continue the question
Related reading
- Bridging Scales: Spectral Theory Reveals How Local Connectivity Rules Sculpt Global Neural Dynamics in Spatially Extended NetworksNeurIPS 2025 · 2025
- Redefining spiking neural networks through the lens of dynamical superspaceCognitive Neurodynamics · 2026
Cite this work
Wentao Wang, Keren Gao, Guozhang Chen. Global Credit Assignment via Dynamical Criticality. ICML 2026.