Research map / Label-efficient and robust learning / Domain adaptation and generalization
The Global Anchor Method for Quantifying Linguistic Shifts and Domain Adaptation
Zi Yin, Vin Sachidananda, Balaji Prabhakar
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 |
|---|---|---|---|
| Use the global anchor distance instead of alignment to quantify corpus-level shifts | Word embeddings trained on separate corpora · Google Books, all pairs of years 1900–2000 | corpus-level shift detection: 0.826 | No clear effect |
| Compare embeddings of different dimensionalities using the global anchor method | Global anchor method · Embeddings with different dimensionalities | applicability | Improved |
| Use matrix multiplications and parallel computation instead of constrained optimization | Global anchor method versus alignment · Corpus-level shift measurement | implementation simplicity | Improved |
| Use distributional global anchors instead of frequency-based comparisons | Language evolution analysis · Google Books English fiction over time | resolution of linguistic shifts | Improved |
| Apply graph Laplacian embedding to global anchor distances | Global anchor method · Google Books English fiction over time | recovery of language-evolution trajectory | Improved |
| Apply graph Laplacian embedding to global anchor distances | Global anchor method · arXiv academic categories | domain clustering | Improved |
| Vary embedding dimensionality | Alignment/global anchor distance ratio · Embedding dimensionalities 100, 200, 400, and 500 | alignment/anchor distance ratio: 0.797, 0.811, 0.814, 0.809 | No clear effect |
Closest papers by meaning
- Fake it Till You Make it: Self-Supervised Semantic Shifts for Monolingual Word Embedding Tasks (AAAI 2021 · similarity 0.89)
- Back to the Future – Temporal Adaptation of Text Representations (AAAI 2020 · similarity 0.86)
- Anchored Diffusion Language Model (NeurIPS 2025 · similarity 0.85)
- Boosting Multi-Domain Fine-Tuning of Large Language Models through Evolving Interactions between Samples (ICML 2025 · similarity 0.85)
- Aligner: Efficient Alignment by Learning to Correct (NeurIPS 2024 · similarity 0.84)
- Diff-eRank: A Novel Rank-Based Metric for Evaluating Large Language Models (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.
