mlsanityData to 14 Oct 2026

Bootstrapping Parallel Anchors for Relative Representations

arXiv 2303.00721 ↗First posted 1 Mar 2023cs.LG cs.AIcited by 8

Not on the map. Why?

Abstract (arXiv)

The use of relative representations for latent embeddings has shown potential in enabling latent space communication and zero-shot model stitching across a wide range of applications. Nevertheless, relative representations rely on a certain amount of parallel anchors to be given as input, which can be impractical to obtain in certain scenarios. To overcome this limitation, we propose an optimization-based method to discover new parallel anchors from a limited known set (seed). Our approach can be used to find semantic correspondence between different domains, align their relative spaces, and achieve competitive results in several tasks.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 24 of its 27 references that are in the map. First posted on arXiv 1 Mar 2023.

Builds on

24 of 27 references.

References · 24

  1. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  2. 2013Efficient Estimation of Word Representations in Vector Spacecited by 4,375
  3. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
  4. 2019Similarity of Neural Network Representations Revisitedcited by 1,755
  5. 2016Enriching Word Vectors with Subword Informationcited by 1,583
  6. 2017Revisiting Unreasonable Effectiveness of Data in Deep Learning Eracited by 919
  7. 2017Word Translation Without Parallel Datacited by 487
  8. 2018Insights on representational similarity in neural networks with canonical correlationcited by 324
  9. 2014Understanding image representations by measuring their equivariance and equivalencecited by 243
  10. 2015Convergent Learning: Do different neural networks learn the same representations?cited by 220
  11. 2021Revisiting Model Stitching to Compare Neural Representationscited by 209
  12. 2013Exploiting Similarities among Languages for Machine Translationcited by 191
12 more
  1. 2022Relative representations enable zero-shot latent space communicationcited by 191
  2. 2020The Multilingual Amazon Reviews Corpuscited by 143
  3. 2019Trivializations for Gradient-Based Optimization on Manifoldscited by 79
  4. 2021Similarity and Matching of Neural Network Representationscited by 78
  5. 2021Representation Topology Divergence: A Method for Comparing Neural Network Representationscited by 66
  6. 2022ASIF: Coupled Data Turns Unimodal Models to Multimodal Without Trainingcited by 47
  7. 2019The Shape of Data: Intrinsic Distance for Data Distributionscited by 38
  8. 2020Are All Good Word Vector Spaces Isomorphic?cited by 37
  9. 2021Low-Dimensional Structure in the Space of Language Representations is Reflected in Brain Responsescited by 34
  10. 2020Fantastic Embeddings and How to Align Them: Zero-Shot Inference in a Multi-Shop Scenariocited by 16
  11. 2021CoReS: Compatible Representations via Stationaritycited by 14
  12. 2022How do Variational Autoencoders Learn? Insights from Representational Similaritycited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 1; 2025: 2; 2026: 4.

8 papers.

  1. 2025Harnessing the Universal Geometry of Embeddingscited by 55
  2. 2023From Bricks to Bridges: Product of Invariances to Enhance Latent Space Communicationcited by 21
  3. 2025Escaping Plato's Cave: Towards the Alignment of 3D and Text Latent Spacescited by 6
  4. 2024TOAST: Transformer Optimization using Adaptive and Simple Transformationscited by 3
  5. 2026Vector Linking via Cross-Model Local Isometric Consistencycited by 1
  6. 2026Relative Geometry of Neural Forecasters: Linking Accuracy and Alignment in Learned Latent Geometrycited by 0
  7. 2026A Stitch in Time Saves Nine: Preserving Policy Compatibility Under Perception Updates in End-to-End Autonomous Drivingcited by 0
  8. 2026Representation Alignment from a Few Paired Data via Distilled Optimal Transportcited by 0