# Less or More From Teacher: Exploiting Trilateral Geometry For Knowledge Distillation

arXiv: https://arxiv.org/abs/2312.15112 · HTML: https://mlsanity.com/paper/2312.15112/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in General knowledge distillation methods, one of 12 directions in Knowledge distillation.
- Builds on 23 of its 31 references that are in the map.
- First posted on arXiv 22 Dec 2023.

## Placed in

- Knowledge distillation / [General knowledge distillation methods](https://mlsanity.com/distill/general-knowledge-distillation-methods/)

## Builds on (most cited first, 23 of 23)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Learning Deep Features for Discriminative Localization](https://mlsanity.com/paper/1512.04150/) (2015)
- [A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks](https://mlsanity.com/paper/1610.02136/) (2016)
- [FitNets: Hints for Thin Deep Nets](https://mlsanity.com/paper/1412.6550/) (2014)
- [A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks](https://mlsanity.com/paper/1807.03888/) (2018)
- [Moment Matching for Multi-Source Domain Adaptation](https://mlsanity.com/paper/1812.01754/) (2018)
- [Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks](https://mlsanity.com/paper/1706.02690/) (2017)
- [Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer](https://mlsanity.com/paper/1612.03928/) (2016)
- [Relational Knowledge Distillation](https://mlsanity.com/paper/1904.05068/) (2019)
- [Contrastive Representation Distillation](https://mlsanity.com/paper/1910.10699/) (2019)
- [DeepFM: A Factorization-Machine based Neural Network for CTR Prediction](https://mlsanity.com/paper/1703.04247/) (2017)
- [Bilevel Programming for Hyperparameter Optimization and Meta-Learning](https://mlsanity.com/paper/1806.04910/) (2018)
- [Knowledge Transfer via Distillation of Activation Boundaries Formed by Hidden Neurons](https://mlsanity.com/paper/1811.03233/) (2018)
- [Knowledge Distillation from A Stronger Teacher](https://mlsanity.com/paper/2205.10536/) (2022)
- [Out-of-Distribution Detection Using an Ensemble of Self Supervised Leave-out Classifiers](https://mlsanity.com/paper/1809.03576/) (2018)
- [Rethinking Soft Labels for Knowledge Distillation: A Bias-Variance Tradeoff Perspective](https://mlsanity.com/paper/2102.00650/) (2021)
- [BAM! Born-Again Multi-Task Networks for Natural Language Understanding](https://mlsanity.com/paper/1907.04829/) (2019)
- [Teacher-Student Architecture for Knowledge Distillation: A Survey](https://mlsanity.com/paper/2308.04268/) (2023)
- [Deep geometric knowledge distillation with graphs](https://mlsanity.com/paper/1911.03080/) (2019)
- [Teacher's pet: understanding and mitigating biases in distillation](https://mlsanity.com/paper/2106.10494/) (2021)
- [Memorize, Factorize, or be Na"ive: Learning Optimal Feature Interaction Methods for CTR Prediction](https://mlsanity.com/paper/2108.01265/) (2021)

## Built on by (most cited first, 2 of 2)

- [VRM: Knowledge Distillation via Virtual Relation Matching](https://mlsanity.com/paper/2502.20760/) (2025)
- [MoKD: Multi-Task Optimization for Knowledge Distillation](https://mlsanity.com/paper/2505.08170/) (2025)

## Abstract (arXiv)

Knowledge distillation aims to train a compact student network using soft supervision from a larger teacher network and hard supervision from ground truths. However, determining an optimal knowledge fusion ratio that balances these supervisory signals remains challenging. Prior methods generally resort to a constant or heuristic-based fusion ratio, which often falls short of a proper balance. In this study, we introduce a novel adaptive method for learning a sample-wise knowledge fusion ratio, exploiting both the correctness of teacher and student, as well as how well the student mimics the teacher on each sample. Our method naturally leads to the intra-sample trilateral geometric relations among the student prediction (S), teacher prediction (T), and ground truth (G). To counterbalance the impact of outliers, we further extend to the inter-sample relations, incorporating the teacher's global average prediction barT for samples within the same class. A simple neural network then learns the implicit mapping from the intra- and inter-sample relations to an adaptive, sample-wise knowledge fusion ratio in a bilevel-optimization manner. Our approach provides a simple, practical, and adaptable solution for knowledge distillation that can be employed across various architectures and model sizes. Extensive experiments demonstrate consistent improvements over other loss re-weighting methods on image classification, attack detection, and click-through rate prediction.