# Adapting Auxiliary Losses Using Gradient Similarity

arXiv: https://arxiv.org/abs/1812.02224 · HTML: https://mlsanity.com/paper/1812.02224/

## Key facts

- Cited by 100 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Dec 2018.

## Built on by (most cited first, 30 of 100)

- [ImageNet-21K Pretraining for the Masses](https://mlsanity.com/paper/2104.10972/) (2021)
- [Multi-Task Learning as a Bargaining Game](https://mlsanity.com/paper/2202.01017/) (2022)
- [Fishr: Invariant Gradient Variances for Out-of-Distribution Generalization](https://mlsanity.com/paper/2109.02934/) (2021)
- [Multi-Task Reinforcement Learning with Context-based Representations](https://mlsanity.com/paper/2102.06177/) (2021)
- [Continual World: A Robotic Benchmark For Continual Reinforcement Learning](https://mlsanity.com/paper/2105.10919/) (2021)
- [The CLRS Algorithmic Reasoning Benchmark](https://mlsanity.com/paper/2205.15659/) (2022)
- [Evading the Simplicity Bias: Training a Diverse Set of Models Discovers Solutions with Superior OOD Generalization](https://mlsanity.com/paper/2105.05612/) (2021)
- [Auto-Lambda: Disentangling Dynamic Task Relationships](https://mlsanity.com/paper/2202.03091/) (2022)
- [On The Effect of Auxiliary Tasks on Representation Dynamics](https://mlsanity.com/paper/2102.13089/) (2021)
- [ConFIG: Towards Conflict-free Training of Physics Informed Neural Networks](https://mlsanity.com/paper/2408.11104/) (2024)
- [Make Continual Learning Stronger via C-Flat](https://mlsanity.com/paper/2404.00986/) (2024)
- [MetaBalance: Improving Multi-Task Recommendations via Adapting Gradient Magnitudes of Auxiliary Tasks](https://mlsanity.com/paper/2203.06801/) (2022)
- [Gradient Imitation Reinforcement Learning for Low Resource Relation Extraction](https://mlsanity.com/paper/2109.06415/) (2021)
- [Adaptive Transfer Learning on Graph Neural Networks](https://mlsanity.com/paper/2107.08765/) (2021)
- [ForkMerge: Mitigating Negative Transfer in Auxiliary-Task Learning](https://mlsanity.com/paper/2301.12618/) (2023)
- [Avoiding Catastrophe: Active Dendrites Enable Multi-Task Learning in Dynamic Environments](https://mlsanity.com/paper/2201.00042/) (2021)
- [Fast-Slow Test-Time Adaptation for Online Vision-and-Language Navigation](https://mlsanity.com/paper/2311.13209/) (2023)
- [It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher](https://mlsanity.com/paper/2203.17008/) (2022)
- [Scalarization for Multi-Task and Multi-Domain Learning at Scale](https://mlsanity.com/paper/2310.08910/) (2023)
- [Deep AutoAugment](https://mlsanity.com/paper/2203.06172/) (2022)
- [Exploring the Relationship between Alignment and Cross-lingual Transfer in Multilingual Transformers](https://mlsanity.com/paper/2306.02790/) (2023)
- [Auxiliary Task Update Decomposition: The Good, The Bad and The Neutral](https://mlsanity.com/paper/2108.11346/) (2021)
- [A Survey of State Representation Learning for Deep Reinforcement Learning](https://mlsanity.com/paper/2506.17518/) (2025)
- [Mixture-of-Skills: Learning to Optimize Data Usage for Fine-Tuning Large Language Models](https://mlsanity.com/paper/2406.08811/) (2024)
- [Lipschitzness Is All You Need To Tame Off-policy Generative Adversarial Imitation Learning](https://mlsanity.com/paper/2006.16785/) (2020)
- [AMAGO-2: Breaking the Multi-Task Barrier in Meta-Reinforcement Learning with Transformers](https://mlsanity.com/paper/2411.11188/) (2024)
- [Multi-Agent Policy Transfer via Task Relationship Modeling](https://mlsanity.com/paper/2203.04482/) (2022)
- [Improving Few-Shot Generalization by Exploring and Exploiting Auxiliary Data](https://mlsanity.com/paper/2302.00674/) (2023)
- [Auxiliary Learning as an Asymmetric Bargaining Game](https://mlsanity.com/paper/2301.13501/) (2023)
- [Fix-A-Step: Semi-supervised Learning from Uncurated Unlabeled Data](https://mlsanity.com/paper/2208.11870/) (2022)

## Abstract (arXiv)

One approach to deal with the statistical inefficiency of neural networks is to rely on auxiliary losses that help to build useful representations. However, it is not always trivial to know if an auxiliary task will be helpful for the main task and when it could start hurting. We propose to use the cosine similarity between gradients of tasks as an adaptive weight to detect when an auxiliary loss is helpful to the main loss. We show that our approach is guaranteed to converge to critical points of the main task and demonstrate the practical usefulness of the proposed algorithm in a few domains: multi-task supervised learning on subsets of ImageNet, reinforcement learning on gridworld, and reinforcement learning on Atari games.