mlsanityData to 14 Oct 2026

Hitting "Probe"rty with Non-Linearity, and More

arXiv 2402.16168 ↗First posted 25 Feb 2024cs.CL cs.AIcited by 0

On the map: Mechanistic interpretability → Knowledge and representations inside LLMs

Abstract (arXiv)

Structural probes learn a linear transformation to find how dependency trees are embedded in the hidden states of language models. This simple design may not allow for full exploitation of the structure of the encoded information. Hence, to investigate the structure of the encoded information to its full extent, we incorporate non-linear structural probes. We reformulate the design of non-linear structural probes introduced by White et al. making its design simpler yet effective. We also design a visualization framework that lets us qualitatively assess how strongly two words in a sentence are connected in the predicted dependency tree. We use this technique to understand which non-linear probe variant is good at encoding syntactical information. Additionally, we also use it to qualitatively investigate the structure of dependency trees that BERT encodes in each of its layers. We find that the radial basis function (RBF) is an effective non-linear probe for the BERT model than the linear probe.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Knowledge and representations inside LLMs, one of 13 directions in Mechanistic interpretability. Builds on 8 of its 13 references that are in the map. First posted on arXiv 25 Feb 2024.

Builds on

8 of 13 references.

References · 8

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  3. 2020A Primer in BERTology: What we know about how BERT workscited by 815
  4. 2019Visualizing and Measuring the Geometry of BERTcited by 216
  5. 2018Colorless green recurrent networks dream hierarchicallycited by 213
  6. 2019Neural Language Models as Psycholinguistic Subjects: Representations of Syntactic Statecited by 138
  7. 2021A Non-Linear Structural Probecited by 31
  8. 2018Distinct patterns of syntactic agreement errors in recurrent networks and humanscited by 13

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Can We Use Probing to Better Understand Fine-tuning and Knowledge Distillation of the BERT NLU?
  2. 2022Probing for Labeled Dependency Trees
  3. 2026Linear Probes Detect Task Format, Not Reasoning Mode in Language Model Hidden States
  4. 2024Analysis and Visualization of Linguistic Structures in Large Language Models: Neural Representations of Verb-Particle Constructions in BERT
  5. 2023Rethinking the Construction of Effective Metrics for Understanding the Mechanisms of Pretrained Language Models
  6. 2026StructLens: A Structural Lens for Language Models via Maximum Spanning Trees
  7. 2023Interventional Probing in High Dimensions: An NLI Case Study
  8. 2026STRUCTURALCOST: A controlled reading time dataset for modeling human sentence processing difficulty