mlsanityData to 13 Oct 2026

Geographic and Geopolitical Biases of Language Models

arXiv 2212.10408 ↗First posted 20 Dec 2022cs.CLcited by 29

Not on the map. Why?

Abstract (arXiv)

Pretrained language models (PLMs) often fail to fairly represent target users from certain world regions because of the under-representation of those regions in training datasets. With recent PLMs trained on enormous data sources, quantifying their potential biases is difficult, due to their black-box nature and the sheer scale of the data sources. In this work, we devise an approach to study the geographic bias (and knowledge) present in PLMs, proposing a Geographic-Representation Probing Framework adopting a self-conditioning method coupled with entity-country mappings. Our findings suggest PLMs' representations map surprisingly well to the physical world in terms of country-to-country associations, but this knowledge is unequally shared across languages. Last, we explain how large PLMs despite exhibiting notions of geographical proximity, over-amplify geopolitical favouritism at inference time.

Key facts. Cited by 29 papers in the map (data through 13 Oct 2026). Builds on 16 of its 36 references that are in the map. First posted on arXiv 20 Dec 2022.

Builds on

16 of 36 references.

References · 16

  1. 2020Language Models are Few-Shot Learnerscited by 32,320
  2. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,500
  3. 2019Plug and Play Language Models: A Simple Approach to Controlled Text Generationcited by 797
  4. 2021BARTScore: Evaluating Generated Text as Text Generationcited by 720
  5. 2021Towards Understanding and Mitigating Social Biases in Language Modelscited by 296
  6. 2021Generating Datasets with Pretrained Language Modelscited by 193
  7. 2022mGPT: Few-Shot Learners Go Multilingualcited by 141
  8. 2022Crossmodal-3600: A Massively Multilingual Multimodal Evaluation Datasetcited by 100
  9. 2021Mitigating Language-Dependent Ethnic Bias in BERTcited by 87
  10. 2021Towards Zero-Label Language Learningcited by 76
  11. 2022PromDA: Prompt-based Data Augmentation for Low-Resource NLU Taskscited by 51
  12. 2021Detecting Cross-Geographic Biases in Toxicity Modeling on Social Mediacited by 31
4 more
  1. 2022GeoMLAMA: Geo-Diverse Commonsense Probing on Multilingual Pre-Trained Language Modelscited by 30
  2. 2021Dataset Geography: Mapping Language Data to Language Userscited by 22
  3. 2022Using Natural Sentences for Understanding Biases in Language Modelscited by 19
  4. 2021Probing as Quantifying Inductive Biascited by 12

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2023: 4; 2024: 8; 2025: 10; 2026: 7.

29 papers.

Sort:

By area:

  1. 2024Understanding and Mitigating Language Confusion in LLMscited by 75
  2. 2024LLMmap: Fingerprinting For Large Language Modelscited by 58
  3. 2023This Land is {Your, My} Land: Evaluating Geopolitical Biases in Language Modelscited by 31
  4. 2024TorchSpatial: A Location Encoding Framework and Benchmark for Spatial Representation Learningcited by 29
  5. 2023Are Large Language Models Temporally Grounded?cited by 26
  6. 2025BiasEdit: Debiasing Stereotyped Language Models via Model Editingcited by 18
  7. 2023On the Promises and Challenges of Multimodal Foundation Models for Geographical, Environmental, Agricultural, and Urban Planning Applicationscited by 15
  8. 2024What Do Language Models Hear? Probing for Auditory Representations in Language Modelscited by 15
  9. 2024Monotonic Representation of Numeric Properties in Language Modelscited by 15
  10. 2024On the Scaling Laws of Geographical Representation in Language Modelscited by 14
  11. 2025Spatial-RAG: Spatial Retrieval Augmented Generation for Real-World Geospatial Reasoning Questionscited by 13
  12. 2025Steering into New Embedding Spaces: Analyzing Cross-Lingual Alignment Induced by Model Interventions in Multilingual Language Modelscited by 10
17 more
  1. 2023Transformation vs Tradition: Artificial General Intelligence (AGI) for Arts and Humanitiescited by 9
  2. 2025OmniGeo: Towards a Multimodal Large Language Models for Geospatial Artificial Intelligencecited by 7
  3. 2024Evaluation of Geographical Distortions in Language Modelscited by 6
  4. 2024Attention Speaks Volumes: Localizing and Mitigating Bias in Language Modelscited by 6
  5. 2025Whose Truth? Pluralistic Geo-Alignment for (Agentic) AIcited by 6
  6. 2026In Agents We Trust, but Who Do Agents Trust? Latent Source Preferences Steer LLM Generationscited by 4
  7. 2025FLUKE: A Linguistically-Driven and Task-Agnostic Framework for Robustness Evaluationcited by 3
  8. 2025GeoBS: Information-Theoretic Quantification of Geographic Bias in AI Modelscited by 3
  9. 2026Regional Bias in Large Language Modelscited by 1
  10. 2025ExpertLens: Activation steering features are highly interpretablecited by 0
  11. 2025Improving Social Determinants of Health Documentation in French EHRs Using Large Language Modelscited by 0
  12. 2025Coordinates from Context: Using LLMs to Ground Complex Location Referencescited by 0
  13. 2026IndRegBias: A Dataset for Studying Indian Regional Biases in English and Code-Mixed Social Media Commentscited by 0
  14. 2026Per-Entity Bias Mapping for AI Visibility: Why Brand Mentions Require Entity-Specific Calibrationcited by 0
  15. 2026Evaluating Regional Bias in LLMs From Abstract Stereotype to Concrete Social Decision-Makingcited by 0
  16. 2026Mapping the City Through the Lens of Language Modelscited by 0
  17. 2026Wiktionary as a Crowdsourced Lexicon for English Dialectscited by 0