# Technical Report: Quantifying and Analyzing the Generalization Power of a DNN

arXiv: https://arxiv.org/abs/2505.06993 · HTML: https://mlsanity.com/paper/2505.06993/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 31 references that are in the map.
- First posted on arXiv 11 May 2025.

## Builds on (most cited first, 21 of 21)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Sharpness-Aware Minimization for Efficiently Improving Generalization](https://mlsanity.com/paper/2010.01412/) (2020)
- [Visualizing the Loss Landscape of Neural Nets](https://mlsanity.com/paper/1712.09913/) (2017)
- [A Unified Approach to Interpreting and Boosting Adversarial Transferability](https://mlsanity.com/paper/2010.04055/) (2020)
- [Discovering and Explaining the Representation Bottleneck of DNNs](https://mlsanity.com/paper/2111.06236/) (2021)
- [Defining and Quantifying the Emergence of Sparse Concepts in DNNs](https://mlsanity.com/paper/2111.06206/) (2021)
- [Interpreting and Boosting Dropout from a Game-Theoretic View](https://mlsanity.com/paper/2009.11729/) (2020)
- [Does a Neural Network Really Encode Symbolic Concepts?](https://mlsanity.com/paper/2302.13080/) (2023)
- [Explaining Generalization Power of a DNN Using Interactive Concepts](https://mlsanity.com/paper/2302.13091/) (2023)
- [Noise-Robust De-Duplication at Scale](https://mlsanity.com/paper/2210.04261/) (2022)
- [Defining and Extracting generalizable interaction primitives from DNNs](https://mlsanity.com/paper/2401.16318/) (2024)
- [Towards the Dynamics of a DNN Learning Symbolic Interactions](https://mlsanity.com/paper/2407.19198/) (2024)
- [Proving Common Mechanisms Shared by Twelve Methods of Boosting Adversarial Transferability](https://mlsanity.com/paper/2207.11694/) (2022)
- [Can the Inference Logic of Large Language Models be Disentangled into Symbolic Concepts?](https://mlsanity.com/paper/2304.01083/) (2023)
- [Bayesian Neural Networks Avoid Encoding Complex and Perturbation-Sensitive Concepts](https://mlsanity.com/paper/2302.13095/) (2023)
- [Where We Have Arrived in Proving the Emergence of Sparse Symbolic Concepts in AI Models](https://mlsanity.com/paper/2305.01939/) (2023)
- [Two-Phase Dynamics of Interactions Explains the Starting Point of a DNN Learning Over-Fitted Features](https://mlsanity.com/paper/2405.10262/) (2024)
- [Layerwise Change of Knowledge in Neural Networks](https://mlsanity.com/paper/2409.08712/) (2024)

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

- [Cross-LLM Consistency in Inference: Evidence from Shared Interactions](https://mlsanity.com/paper/2606.08129/) (2026)
- [How Does Parameter Pruning Reshape DNN Representations? An Interaction-Driven Exploration](https://mlsanity.com/paper/2609.06483/) (2026)

## Abstract (arXiv)

This paper proposes a new perspective for analyzing the generalization power of deep neural networks (DNNs), i.e., directly disentangling and analyzing the dynamics of generalizable and non-generalizable interaction encoded by a DNN through the training process. Specifically, this work builds upon the recent theoretical achievement in explainble AI, which proves that the detailed inference logic of DNNs can be can be strictly rewritten as a small number of AND-OR interaction patterns. Based on this, we propose an efficient method to quantify the generalization power of each interaction, and we discover a distinct three-phase dynamics of the generalization power of interactions during training. In particular, the early phase of training typically removes noisy and non-generalizable interactions and learns simple and generalizable ones. The second and the third phases tend to capture increasingly complex interactions that are harder to generalize. Experimental results verify that the learning of non-generalizable interactions is the the direct cause for the gap between the training and testing losses.