The DOI matches its Crossref record; 88 of the 88 references it cites were found in a registry or an index; 29 of the 50 sources its claims cite were checked for support.
Generative language models exhibit social identity biases
Open accessAbstract
Social identity biases, particularly the tendency to favor one’s own group (ingroup solidarity) and derogate other groups (outgroup hostility), are deeply rooted in human psychology and social behavior. However, it is unknown if such biases are also present in artificial intelligence systems. Here we show that large language models (LLMs) exhibit patterns of social identity bias, similarly to humans. By administering sentence completion prompts to 77 different LLMs (for instance, ‘We are…’), we demonstrate that nearly all base models and some instruction-tuned and preference-tuned models display clear ingroup favoritism and outgroup derogation. These biases manifest both in controlled experimental settings and in naturalistic human–LLM conversations. However, we find that careful curation of training data and specialized fine-tuning can substantially reduce bias levels. These findings have important implications for developing more equitable artificial intelligence systems and highlight the urgent need to understand how human–LLM interactions might reinforce existing social biases.
Overview
CoChat Summary
TL;DR
Large language models show human-like social identity biases: ingroup sentences are more likely to be positive and outgroup sentences more likely to be negative. These biases appear across many models, are amplified by partisan fine-tuning, and are also present in real-world human–LLM conversations.
Key findings
- Across 77 LLMs, most models exhibited ingroup solidarity and outgroup hostility; of 56 models tested with the default prompt, only four did not show ingroup solidarity and six did not show outgroup hostility. Pooled mixed-effects regressions found ingroup sentences were 93% more likely to be positive and outgroup sentences 115% more likely to be negative.
- Human-written text from four pretraining corpora also showed bias: pooled estimates were 68% more likely positive for ingroup sentences and 70% more likely negative for outgroup sentences. The ingroup solidarity of 44 LLMs was statistically the same as the human average, and 42 models had statistically similar outgroup hostility.
- Fine-tuning GPT-2, BLOOM, and BLOOMZ on US partisan Twitter data increased bias: ingroup sentences were 361% more likely to be positive and outgroup sentences 550% more likely to be negative, compared with 86% and 83% for the same models without fine-tuning. Removing ingroup-positive or outgroup-negative sentences reduced bias, with 0% of both bringing ingroup solidarity to near parity.
- In real-world conversations from WildChat and LMSYS-Chat-1M, LLM ingroup sentences were 80% more likely to be positive and outgroup sentences 57% more likely to be negative; users showed comparable biases, with ingroup sentences 86% more likely to be positive and outgroup sentences 158% more likely to be negative.
Methods
- Study 1 tested 77 base, instruction-tuned, and preference-tuned LLMs by prompting “We are” or “They are” to generate 2,000 sentences per model, filtering for quality/diversity, classifying sentiment with a RoBERTa-based classifier, and fitting logistic regressions controlling for type-to-token ratio and sentence length.
- Study 2 fine-tuned GPT-2, BLOOM, and BLOOMZ on US partisan Twitter data, then varied the proportion of ingroup-positive and outgroup-negative sentences in the training data and re-measured bias with the same sentiment and regression approach.
- Study 3 analyzed real human–LLM conversations from WildChat and LMSYS-Chat-1M, extracting user and model sentences starting with “We are” or “They are” and using mixed-effects logistic regressions to estimate ingroup solidarity and outgroup hostility.
Limitations
- Stated: the operationalization provides only a simplified view of complex social-psychological phenomena and is not meant as a sentence-level classifier of ingroup solidarity or outgroup hostility.
- Stated: the study is English-centric, limiting generalizability to other languages and cultural contexts, and the bias measure is single-turn while real conversations are often dynamic and multi-turn.
- Stated: real-world conversation findings may be influenced by high bias in user queries, and model alignment may be weaker in multi-turn settings; the paper also notes that WildChat/LMSYS users may be non-representative.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures



Verification
Claims the paper makes
15 Valid · 3 Not valid · 4 Not checked · 55 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“However, we also observe that alignment techniques such as instruction fine-tuning and preference-tuning are effective at reducing social identity bias, corroborating previous research59,70.”
§ Generative language models exhibit social identity biases > Discussion
What the sources say
- Not matchedcites [59] Chung, H. W. et al. Scaling instruction-finetuned language models. J. Mach. Learn. Res. 25, 1–53 (2024).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Disputedcites [70] Bai, Y. et al. Training a helpful and harmless assistant with reinforcement learning from human feedback. Preprint at https://arxiv.org/abs/2204.05862 (2022). sourceExperimental — accuracy not yet measured
Section 7.1 (p. 35) explicitly cautions that more positive sentiment toward racial and religious groups does not necessarily mean bias was reduced, and says gender biases remained strongly correlated with those of the underlying models. Section 4.6.2 (p. 23) describes any improvement on BBQ-Lite as slight and uncertain. The paper also says its fine-tuning used RL rather than a supervised instruction-fine-tuning stage (Section 1.3, p. 8), so it does not establish the claimed effectiveness of both techniques. (LLM, conf=0.85)
Checked by a reasoning model — may be wrong.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“We also conducted robustness checks with other sentiment classification tools95–98, which show broad agreement with the RoBERTa results (Supplementary Sections 5 and 6).”
§ Generative language models exhibit social identity biases > Methods > Text generation and processing
What the sources say
- Supportedcites [95] Hutto, C. & Gilbert, E. VADER: a parsimonious rule-based model for sentiment analysis of social media text. Proc. Int. AAAI Conf. Web Social Media 8, 216–225 (2014). source
The Conclusion (section 6) identifies VADER as a sentiment analysis tool, supporting its citation as one of the other tools used. The agreement with RoBERTa is presented as the citing authors’ own robustness-check result, not a finding attributed to VADER. (LLM, conf=0.90)
- Not matchedcites [96] Årup Nielsen, F. A new ANEW: evaluation of a word list for sentiment analysis in microblogs. In Proc. ESWC2011 Workshop on ‘Making Sense of Microposts’: Big Things Come in Small Packages, CEUR Workshop Proceedings Vol. 718 (eds Rowe, M. et al.) 93–98 (2011).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not matchedcites [97] Loria, S. textblob Documentation, release 0.18.0.post0 edn https://readthedocs.org/projects/textblob/downloads/pdf/dev/ (Readthedocs, 2024).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites [98] Potts, C., Wu, Z., Geiger, A. & Kiela, D. DynaSent: A dynamic benchmark for sentiment analysis. Proc. 59th Annu. Meet. Assoc. Comput. Linguist. 11th Int. Jt. Conf. Nat. Lang. Process. Vol. 1: Long Pap. 2388–2404 (2021). source
The citation identifies a sentiment-classification tool, rather than attributing the citing authors’ robustness-check results to DynaSent. Sections 3.1 and 4.1 describe DynaSent’s models as RoBERTa-based three-way sentiment classifiers. The reported broad agreement is the citing authors’ own result. (LLM, conf=0.88)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Language models, on average, become roughly five times more hostile toward a general (non-specific) outgroup after fine-tuning with US partisan social media data, in line with previous work on outgroup hostility on US social media61.”
§ Generative language models exhibit social identity biases > Discussion
What the sources say
- Supportedcites [61] Rathje, S., Van Bavel, J. J. & van der Linden, S. Out-group animosity drives engagement on social media. Proc. Natl Acad. Sci. USA 118, e2024292118 (2021). source
The citation is attached to “previous work on outgroup hostility on US social media,” not to the citing authors’ fivefold language-model result. The source’s Discussion reports that, across Facebook and Twitter posts, language about political out-groups was associated with more sharing. It does not report fine-tuning language models. (LLM, conf=0.91)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“This signals an asymmetric effect, where fine-tuning with partisan social media data has an especially pronounced effect on outgroup hostility, in line with previous research on the viral potential of outgroup language61,62.”
§ Generative language models exhibit social identity biases > Results > Study 2—training data effects on social identity biases
What the sources say
- Supportedcites [61] Rathje, S., Van Bavel, J. J. & van der Linden, S. Out-group animosity drives engagement on social media. Proc. Natl Acad. Sci. USA 118, e2024292118 (2021). source
The citation is attached to the claim about the viral potential of out-group language, not to the citing authors’ fine-tuning result. The Discussion reports that posts about political out-groups were more likely to be shared and that out-group language was a particularly strong predictor of sharing. Page numbers are unavailable in our copy. (LLM, conf=0.92)
- Not checked — no full text of the sourcecites [62] Abramowitz, A. I. & Webster, S. The rise of negative partisanship and the nationalization of US elections in the 21st century. Elect. Stud. 41, 12–22 (2016). source
hedgedstates a causen = 24,000- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Our results also support previous findings that language models can acquire political bias through fine-tuning73.”
§ Generative language models exhibit social identity biases > Discussion
What the sources say
- Partly supportedcites [73] Feng, S., Park, C. Y., Liu, Y. & Tsvetkov, Y. From pretraining data to language models to downstream tasks: tracking the trails of political biases leading to unfair NLP models. In Proc. 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) 11737–11762 (ACL, 2023); https://aclanthology.org/2023.acl-long.656 sourceExperimental — accuracy not yet measured
In “The Effect of Pretraining with Partisan Corpora” (page unavailable), the paper reports that language models acquire political bias after further pretraining on partisan corpora. Section 2.2 (page unavailable) uses “fine-tune” for a separate downstream-task step; it does not establish that the models acquired political bias through that fine-tuning. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
22 more claims · 1 with their cited sources checked
Create a free account to see all 77 claims and to verify the 1 reference still open.
- Probably not a substantial claim. 24
- It may state two separate claims as one. 36
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“While most research has focused on investigating and addressing bias in one-to-one interactions between humans and LLMs (18–20), less attention has been given to how these biases evolve through repeated communications in populations of LLM agents and, ultimately, in mixed human-LLM ecosystems (15), even though the safety of a single LLM does not necessarily imply the safety of a multi-agent system”
From Emergent social conventions and collective bias in LLM populations (2025) · cites this paper as [20]
Experimental — accuracy not yet measuredIn the citing paper
“Recent studies have explored whether LLMs replicate human heuristics (4, 42, 43) or reflect ideological biases learned during training (44–46), but most focus on output-level metrics such as accuracy or bias. Few examine how these outputs are produced.”
From The simulation of judgment in LLMs (2025) · cites this paper as [42]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Generative language models exhibit social identity biases > Results > Study 2—training data effects on social identity biases
“The ingroup solidarity and outgroup hostility produced by the resulting models are depicted in Fig. 2.”
§ Generative language models exhibit social identity biases > Results > Study 2—training data effects on social identity biases
“Fig. 3 Study 2—ingroup solidarity and outgroup hostility of Republican and Democrat fine-tuned models after removing different proportions of positive and negative ingroup and outgroup sentences from training data.”
Citations
Cites
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515 · cited 1×
“Outgroup Hostility Ingroup Solidarity”
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929
Scaling Laws for Neural Language Models
Kaplan, J., McCandlish, S., Henighan, T., Brown, T. B., Chess, B., Child, R., Gray, S., Radford, A., Wu, J., & Amodei, D. (2020) · arXiv · doi:10.48550/arxiv.2001.08361 · Cited by 9,550 · cited 1×
“0045) but not ingroup solidarity (see SI).”
Transformers: State-of-the-Art Natural Language Processing
Wolf, T., Debut, L., Sanh, V., Chaumond, J., Delangue, C., Moi, A., Cistac, P., Rault, T., Louf, R., Funtowicz, M., et al. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations · doi:10.18653/v1/2020.emnlp-demos.6 · Cited by 8,741
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988 · cited 2×
“GPT-2-Large Positive .5106 .9375 2 Results”
VADER: a parsimonious rule-based model for sentiment analysis of social media text
Hutto, C., & Gilbert, E. (2014) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v8i1.14550 · Cited by 6,087
Social categorization and intergroup behaviour
Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971) · European Journal of Social Psychology · doi:10.1002/ejsp.2420010202 · Cited by 5,562 · cited 1×
“Social psychologists have shown that even arbitrary distinctions can lead to immediate intergroup discrimination [12, 13].”
LLaMA: Open and Efficient Foundation Language Models
Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.-A., Lacroix, T., Rozière, B., Goyal, N., Hambro, E., Azhar, F., et al. (2023) · arXiv · doi:10.48550/arxiv.2302.13971 · Cited by 4,602
The Origins and Consequences of Affective Polarization in the United States
Iyengar, S., Lelkes, Y., Levendusky, M., Malhotra, N., & Westwood, S. J. (2019) · Annual Review of Political Science · doi:10.1146/annurev-polisci-051117-073034 · Cited by 3,603 · cited 3×
“This approach is similar to the “feeling thermometer" measure frequently utilized by political scientists and psychologists in assessing affective polarization [9].”
Research works printed but not linked: 12
- Hogg, M. A. & Abrams, D. Social Identifications: A Social Psychology of Intergroup Relations and Group Processes (Taylor & Francis, 1988).
- Santurkar, S. et al. Whose opinions do language models reflect? In Proc. 40th International Conference on Machine Learning 1244, 29971–30004 (ACM, 2023).
- Zhao, W. et al. WildChat: 1M ChatGPT interaction logs in the wild. In Proc. Twelfth International Conference on Learning Representations (ICLR, 2024); https://openreview.net/forum?id=Bl8u7ZRlbM
- Jiang, H., Beeferman, D., Roy, B. & Roy, D. CommunityLM: probing partisan worldviews from language models. In Proc. 29th International Conference on Computational Linguistics 6818–6826 (ACL, 2022); https://aclanthology.org/2022.coling-1.593
- Bolukbasi, T., Chang, K.-W., Zou, J. Y., Saligrama, V. & Kalai, A. T. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. In Advances in Neural Information Processing Systems Vol. 29 (Curran Associates, 2016); https://proceedings.neurips.cc/paper_files/paper/2016/hash/a486cd07e4ac3d270571622f4f316ec5-Abstract.html
- Liang, P. P., Wu, C., Morency, L.-P. & Salakhutdinov, R. Towards understanding and mitigating social biases in language models. In Proc. 38th International Conference on Machine Learning 139 (eds Meila, M. & Zhang, T.) 6565–6576 (PMLR, 2021); https://proceedings.mlr.press/v139/liang21a.html
- Sharma, M. et al. Towards understanding sycophancy in language models. In Proc. Twelfth International Conference on Learning Representations (ICLR, 2024); https://openreview.net/forum?id=tvhaxkMKAn
- Tay, Y. et al. UL2: unifying language learning paradigms. In The Eleventh International Conference on Learning Representations (ICLR, 2023); https://openreview.net/forum?id=6ruVLB727MC
- Zhu, B. et al. Starling-7B: improving helpfulness and harmlessness with RLAIF. In Proc. First Conference on Language Modeling https://openreview.net/forum?id=GqDntYTTbk (2024).
- Holtzman, A., Buys, J., Du, L., Forbes, M. & Choi, Y. The curious case of neural text degeneration. In Proc. International Conference on Learning Representations (ICLR, 2020); https://openreview.net/forum?id=rygGQyrFvH
- Årup Nielsen, F. A new ANEW: evaluation of a word list for sentiment analysis in microblogs. In Proc. ESWC2011 Workshop on ‘Making Sense of Microposts’: Big Things Come in Small Packages, CEUR Workshop Proceedings Vol. 718 (eds Rowe, M. et al.) 93–98 (2011).
- Boyd, R. L., Ashokkumar, A., Seraj, S. & Pennebaker, J. W. The Development and Psychometric Properties of LIWC-22. (Univ. Texas at Austin, 2022).
Cited by
GPT is an effective tool for multilingual psychological text analysis
Rathje, S., Mirea, D.-M., Sucholutsky, I., Marjieh, R., Robertson, C. E., & Van Bavel, J. J. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2308950121 · Cited by 297
“Other work has found that GPT is biased toward responses on cognitive tasks that are similar to those of WEIRD populations (78).”
Self-citationUsing natural language processing to analyse text data in behavioural science
Feuerriegel, S., Maarouf, A., Bär, D., Geissler, D., Schweisthal, J., Pröllochs, N., Robertson, C. E., Rathje, S., Hartmann, J., Mohammad, S. M., et al. (2025) · Nature Reviews Psychology · doi:10.1038/s44159-024-00392-z · Cited by 140
Self-citationGPT is an effective tool for multilingual psychological text analysisPreprint
Rathje, S., Mirea, D.-M., Sucholutsky, I., Marjieh, R., Robertson, C., & Bavel, J. J. V. (2023) · doi:10.31234/osf.io/sekf5 · Cited by 129
Self-citationEmergent social conventions and collective bias in LLM populations
Ashery, A. F., Aiello, L. M., & Baronchelli, A. (2025) · Science Advances · doi:10.1126/sciadv.adu9368 · Cited by 81
“While most research has focused on investigating and addressing bias in one- to-one interactions between humans and LLMs ( 18 – 20 ), less attention has been given to how these biases evolve through repeated communications in populations of LLM agents and, ultimately, in mixed human-LLM ecosystems…”
The simulation of judgment in LLMs
Loru, E., Nudo, J., Di Marco, N., Santirocchi, A., Atzeni, R., Cinelli, M., Cestari, V., Rossi-Arnaud, C., & Quattrociocchi, W. (2025) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2518443122 · Cited by 57
“Recent studies have explored whether LLMs replicate human heuristics [4, 42, 43] or reflect ideological biases learned during training [44–46], but most focus on output-level metrics such as accuracy or bias.”
Metadata
- Authors
- Hu, T., Kyrychenko, Y., Rathje, S., Collier, N., van der Linden, S., & Roozenbeek, J.
- Affiliations
- University of Cambridge (GB)
- New York University (US)
- King's College London (GB)
- Corresponding author
- Hu, Tiancheng
- Kyrychenko, Yara
- Venue
- Nature Computational Science
- Published
- 2024-12-12
- Volume
- 5
- Issue
- 1
- Pages
- 65-75
- ISSN
- 2662-8457 (ISSN-L)
- PMC
- PMC11774750
- OpenAlex
- W4405298817
- Cited by
- 100
- References
- 75
- Citations by year
- 2026: 57 · 2025: 42
- FWCI
- 40.87
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Nature Computational Science (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion, other-oa)
- Apollo (University of Cambridge) (submittedVersion, cc-by)
- arXiv (Cornell University)
- Funders
- Bill and Melinda Gates Foundation (OPP114)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Sociology and Political Science › Social and Intergroup Psychology
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- in-group favoritism, outgroup derogation, large language models, algorithmic bias, human-LLM interaction, training data curation, preference tuning, instruction tuning, sentence completion
- MeSH terms
- Artificial Intelligence, Humans, Language, Social Behavior, Social Identification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-23, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-23 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Generative language models exhibit social identity biases. https://app.cochat.ai/papers/10.1038/s43588-024-00741-1
- The paper
- Hu, T., Kyrychenko, Y., Rathje, S., Collier, N., van der Linden, S., & Roozenbeek, J. (2024). Generative language models exhibit social identity biases. Nature Computational Science. https://doi.org/10.1038/s43588-024-00741-1
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.