# DirectProbe: Studying Representations without Classifiers

arXiv: https://arxiv.org/abs/2104.05904 · HTML: https://mlsanity.com/paper/2104.05904/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Builds on 43 of its 61 references that are in the map.
- First posted on arXiv 13 Apr 2021.

## Builds on (most cited first, 30 of 43)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Deep contextualized word representations](https://mlsanity.com/paper/1802.05365/) (2018)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [BERT Rediscovers the Classical NLP Pipeline](https://mlsanity.com/paper/1905.05950/) (2019)
- [Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned](https://mlsanity.com/paper/1905.09418/) (2019)
- [Are Sixteen Heads Really Better than One?](https://mlsanity.com/paper/1905.10650/) (2019)
- [How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings](https://mlsanity.com/paper/1909.00512/) (2019)
- [Designing and Interpreting Probes with Control Tasks](https://mlsanity.com/paper/1909.03368/) (2019)
- [What you can cram into a single vector: Probing sentence embeddings for linguistic properties](https://mlsanity.com/paper/1805.01070/) (2018)
- [Linguistic Knowledge and Transferability of Contextual Representations](https://mlsanity.com/paper/1903.08855/) (2019)
- [Revealing the Dark Secrets of BERT](https://mlsanity.com/paper/1908.08593/) (2019)
- [Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Fly](https://mlsanity.com/paper/1911.03343/) (2019)
- [Targeted Syntactic Evaluation of Language Models](https://mlsanity.com/paper/1808.09031/) (2018)
- [Information-Theoretic Probing with Minimum Description Length](https://mlsanity.com/paper/2003.12298/) (2020)
- [Information-Theoretic Probing for Linguistic Structure](https://mlsanity.com/paper/2004.03061/) (2020)
- [To Tune or Not to Tune? Adapting Pretrained Representations to Diverse Tasks](https://mlsanity.com/paper/1903.05987/) (2019)
- [oLMpics -- On what Language Model Pre-training Captures](https://mlsanity.com/paper/1912.13283/) (2019)
- [Unsupervised Domain Clusters in Pretrained Language Models](https://mlsanity.com/paper/2004.02105/) (2020)
- [Do NLP Models Know Numbers? Probing Numeracy in Embeddings](https://mlsanity.com/paper/1909.07940/) (2019)
- [Probing the Probing Paradigm: Does Probing Accuracy Entail Task Relevance?](https://mlsanity.com/paper/2005.00719/) (2020)
- [Intermediate-Task Transfer Learning with Pretrained Models for Natural Language Understanding: When and Why Does It Work?](https://mlsanity.com/paper/2005.00628/) (2020)
- [What Happens To BERT Embeddings During Fine-tuning?](https://mlsanity.com/paper/2004.14448/) (2020)
- [Perturbed Masking: Parameter-free Probing for Analyzing and Interpreting BERT](https://mlsanity.com/paper/2004.14786/) (2020)
- [Classification and Clustering of Arguments with Contextualized Word Embeddings](https://mlsanity.com/paper/1906.09821/) (2019)
- [Pareto Probing: Trading Off Accuracy for Complexity](https://mlsanity.com/paper/2010.02180/) (2020)
- [Predicting Performance for Natural Language Processing Tasks](https://mlsanity.com/paper/2005.00870/) (2020)
- [Probing What Different NLP Tasks Teach Machines about Function Word Comprehension](https://mlsanity.com/paper/1904.11544/) (2019)

## Built on by (most cited first, 25 of 25)

- [CREPE: Can Vision-Language Foundation Models Reason Compositionally?](https://mlsanity.com/paper/2212.07796/) (2022)
- [A Closer Look at How Fine-tuning Changes BERT](https://mlsanity.com/paper/2106.14282/) (2021)
- [How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy](https://mlsanity.com/paper/2109.04740/) (2021)
- [The Life Cycle of Knowledge in Big Language Models: A Survey](https://mlsanity.com/paper/2303.07616/) (2023)
- [Probing Graph Representations](https://mlsanity.com/paper/2303.03951/) (2023)
- [Putting Words in BERT's Mouth: Navigating Contextualized Vector Spaces with Pseudowords](https://mlsanity.com/paper/2109.11491/) (2021)
- [Probing via Prompting](https://mlsanity.com/paper/2207.01736/) (2022)
- [Evaluating Contextualized Representations of (Spanish) Ambiguous Words: A New Lexical Resource and Empirical Analysis](https://mlsanity.com/paper/2406.14678/) (2024)
- [On the data requirements of probing](https://mlsanity.com/paper/2202.12801/) (2022)
- [A Critical Study of What Code-LLMs (Do Not) Learn](https://mlsanity.com/paper/2406.11930/) (2024)
- [CUE: An Uncertainty Interpretation Framework for Text Classifiers Built on Pre-Trained Language Models](https://mlsanity.com/paper/2306.03598/) (2023)
- [On Entity Identification in Language Models](https://mlsanity.com/paper/2506.02701/) (2025)
- [Towards Transparent AI: A Survey on Explainable Language Models](https://mlsanity.com/paper/2509.21631/) (2025)
- [An Empirical Study on the Transferability of Transformer Modules in Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2302.00378/) (2023)
- [Analyzing Text Representations by Measuring Task Alignment](https://mlsanity.com/paper/2305.19747/) (2023)
- [Is Probing All You Need? Indicator Tasks as an Alternative to Probing Embedding Spaces](https://mlsanity.com/paper/2310.15905/) (2023)
- [CAST: Compositional Analysis via Spectral Tracking for Understanding Transformer Layer Functions](https://mlsanity.com/paper/2510.14262/) (2025)
- [On the Geometric Structure of Layer Updates in Deep Language Models](https://mlsanity.com/paper/2604.02459/) (2026)
- [Probing Representations for Document-level Event Extraction](https://mlsanity.com/paper/2310.15316/) (2023)
- [Investigating the Representation of Backchannels and Fillers in Fine-tuned Language Models](https://mlsanity.com/paper/2509.20237/) (2025)
- [When Annotators Disagree, Topology Explains: Mapper, a Topological Tool for Exploring Text Embedding Geometry and Ambiguity](https://mlsanity.com/paper/2510.17548/) (2025)
- [Towards Understanding What State Space Models Learn About Code](https://mlsanity.com/paper/2602.06774/) (2026)
- [Domain Pre-training Impact on Representations](https://mlsanity.com/paper/2505.24455/) (2025)
- [On the Performance of Concept Probing: The Influence of the Data (Extended Version)](https://mlsanity.com/paper/2507.18550/) (2025)
- [Concept Probing: Where to Find Human-Defined Concepts (Extended Version)](https://mlsanity.com/paper/2507.18681/) (2025)

## Abstract (arXiv)

Understanding how linguistic structures are encoded in contextualized embedding could help explain their impressive performance across NLP@. Existing approaches for probing them usually call for training classifiers and use the accuracy, mutual information, or complexity as a proxy for the representation's goodness. In this work, we argue that doing so can be unreliable because different representations may need different classifiers. We develop a heuristic, DirectProbe, that directly studies the geometry of a representation by building upon the notion of a version space for a task. Experiments with several linguistic tasks and contextualized embeddings show that, even without training classifiers, DirectProbe can shine light into how an embedding space represents labels, and also anticipate classifier performance for the representation.