Research map / Architectures and efficiency / RNNs and normalization
On the Robustness of Text Vectorizers
Rémi Catellier, Samuel Vaiter, Damien Garreau
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace up to |S| tokens in a document | Concatenation · Documents of fixed length | embedding distance: 1/2-Hölder | Improved |
| Replace up to |S| tokens in a document | Non-normalized TF-IDF · Documents of fixed length | embedding distance: 4m_{\max} v_{\max} \frac{\card{\mathcal{S}}}{T} | Improved |
| Replace up to |S| tokens in a document | Normalized TF-IDF · Documents of fixed length | embedding distance: \frac{4m_{\max}^{1/2}v_{\max}^{1/2}D^{1/4}}{v_{\min}^{1/2}} \sqrt{\frac{\card{\mathcal{S}}}{T}} | Improved |
| Replace a sufficiently small subset of tokens | doc2vec · Documents of fixed length | embedding distance: L \frac{\card{\mathcal{S}}}{T} | Improved |
| Keep the number of replacements fixed while increasing document length | doc2vec · IMDB movie reviews | embedding distance: 1/T | Improved |
Closest papers by meaning
- AdaptDel: Adaptable Deletion Rate Randomized Smoothing for Certified Robustness (NeurIPS 2025 · similarity 0.85)
- Certified Robustness Under Bounded Levenshtein Distance (ICLR 2025 · similarity 0.85)
- Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Models (ICLR 2024 · similarity 0.84)
- Remarkable Robustness of LLMs: Stages of Inference? (NeurIPS 2025 · similarity 0.84)
- The Foundations of Tokenization: Statistical and Computational Concerns (ICLR 2025 · similarity 0.84)
- On the Noise Robustness of In-Context Learning for Text Generation (NeurIPS 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
