paper-with-me

홈 › Papers

Using Degeneracy in the Loss Landscape for Mechanistic Interpretability

2024-05-17 · Lucius Bushnaq, Jake Mendel, Stefan Heimersheim, Dan Braun, Nicholas Goldowsky-Dill, Kaarel Hänni, Cindy Wu, Marius Hobbhahn

Mechanistic Interpretability aims to reverse engineer the algorithms implemented by neural networks by studying their weights and activations. An obstacle to reverse engineering neural networks is that many of the parameters inside a network are not involved in the computation being implemented by the network. These degenerate parameters may obfuscate internal structure. Singular learning theory teaches us that neural network parameterizations are biased towards being more degenerate, and parameterizations with more degeneracy are likely to generalize further. We identify 3 ways that network parameters can be degenerate: linear dependence between activations in a layer; linear dependence between gradients passed back to a layer; ReLUs which fire on the same subset of datapoints. We also present a heuristic argument that modular networks are likely to be more degenerate, and we develop a metric for identifying modules in a network that is based on this argument. We propose that if we can represent a neural network in a way that is invariant to reparameterizations that exploit the degeneracies, then this representation is likely to be more interpretable, and we provide some evidence that such a representation is likely to have sparser interactions. We introduce the Interaction Basis, a tractable technique to obtain a representation that is invariant to degeneracies from linear dependence of activations or Jacobians.

📄 PDF Abstract BibTeX arXiv:2405.10927

Code (0)

등록된 구현이 없습니다.

Tasks

Learning Theory

Similar Papers 제목 키워드 기반

From Global to Local: A Scalable Benchmark for Local Posterior Sampling

2025-07-29 · Rohan Hitchcock, Jesse Hoogland arxiv

Degeneracy is an inherent feature of the loss landscape of neural networks, but it is not well understood how stochastic gradient MCMC (SGMCMC) algorithms interact with this degeneracy. In particular, current global conv…

Embedding Principle of Loss Landscape of Deep Neural Networks

2021-05-30 · NeurIPS 2021 12 · Yaoyu Zhang, Zhongwang Zhang, Tao Luo, Zhi-Qin John Xu

Understanding the structure of loss landscape of deep neural networks (DNNs)is obviously important. In this work, we prove an embedding principle that the loss landscape of a DNN "contains" all the critical points of all…

Protein Folding

DLM-Scope: Mechanistic Interpretability of Diffusion Language Models via Sparse Autoencoders

2026-02-05 · Xu Wang, Bingqing Jiang, Yu Wan, Baosong Yang 외 arxiv

Sparse autoencoders (SAEs) have become a standard tool for mechanistic interpretability in autoregressive large language models (LLMs), enabling researchers to extract sparse, human-interpretable features and intervene o…

Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networks

2025-11-24 · Bianka Kowalska, Halina Kwaśnicka arxiv

The black box nature of deep neural networks poses a significant challenge for the deployment of transparent and trustworthy artificial intelligence (AI) systems. With the growing presence of AI in society, it becomes in…

Prisma: An Open Source Toolkit for Mechanistic Interpretability in Vision and Video

2025-04-28 · Sonia Joseph, Praneet Suresh, Lorenz Hufe, Edward Stevinson 외

Robust tooling and publicly available pre-trained models have helped drive recent advances in mechanistic interpretability for language models. However, similar progress in vision mechanistic interpretability has been hi…