# Accelerating Training with Neuron Interaction and Nowcasting Networks

arXiv: https://arxiv.org/abs/2409.04434 · HTML: https://mlsanity.com/paper/2409.04434/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Sep 2024.

## Built on by (most cited first, 12 of 12)

- [On Linear Mode Connectivity of Mixture-of-Experts Architectures](https://mlsanity.com/paper/2509.11348/) (2025)
- [Celo: Training Versatile Learned Optimizers on a Compute Diet](https://mlsanity.com/paper/2501.12670/) (2025)
- [Learning Model Representations Using Publicly Available Model Hubs](https://mlsanity.com/paper/2510.02096/) (2025)
- [The Impact of Model Zoo Size and Composition on Weight Space Learning](https://mlsanity.com/paper/2504.10141/) (2025)
- [Celo2: Towards Learned Optimization Free Lunch](https://mlsanity.com/paper/2602.19142/) (2026)
- [Quasi-Equivariant Metanetworks](https://mlsanity.com/paper/2604.23720/) (2026)
- [Gradient Flow Matching for Learning Update Dynamics in Neural Network Training](https://mlsanity.com/paper/2505.20221/) (2025)
- [Learning from Oblivion: Predicting Knowledge Overflowed Weights via Retrodiction of Forgetting](https://mlsanity.com/paper/2508.05059/) (2025)
- [Leap+Verify: Regime-Adaptive Speculative Weight Prediction for Accelerating Neural Network Training](https://mlsanity.com/paper/2602.19580/) (2026)
- [WeightCLIP: Aligning Datasets and Models for Weight Space Learning](https://mlsanity.com/paper/2607.03551/) (2026)
- [Measuring Structured Predictability in Neural Training Dynamics: A Cross-Regime Study](https://mlsanity.com/paper/2608.15483/) (2026)
- [Permutation-Equivariant Flow Matching for Alignment-Free Neural Weight Generation](https://mlsanity.com/paper/2609.32833/) (2026)

## Abstract (arXiv)

Neural network training can be accelerated when a learnable update rule is used in lieu of classic adaptive optimizers (e.g. Adam). However, learnable update rules can be costly and unstable to train and use. Recently, Jang et al. (2023) proposed a simpler approach to accelerate training based on weight nowcaster networks (WNNs). In their approach, Adam is used for most of the optimization steps and periodically, only every few steps, a WNN nowcasts (predicts near future) parameters. We improve WNNs by proposing neuron interaction and nowcasting (NiNo) networks. In contrast to WNNs, NiNo leverages neuron connectivity and graph neural networks to more accurately nowcast parameters. We further show that in some networks, such as Transformers, modeling neuron connectivity accurately is challenging. We address this and other limitations, which allows NiNo to accelerate Adam training by up to 50% in vision and language tasks.