Papers › Evaluating Large Language Model Biases in Persona-Steered Generation

Evaluating Large Language Model Biases in Persona-Steered Generation

30 May 2024arXiv:2405.20253archive 2025-07-28

Andy Liu, Mona Diab, Daniel Fried

The task of persona-steered text generation requires large language models (LLMs) to generate text that reflects the distribution of views that an individual fitting a persona could have. People have multifaceted personas, but prior work on bias in LLM-generated opinions has only explored multiple-choice settings or one-dimensional personas. We define an incongruous persona as a persona with multiple traits where one trait makes its other traits less likely in human survey data, e.g. political liberals who support increased military spending. We find that LLMs are 9.7% less steerable towards incongruous personas than congruous ones, sometimes generating the stereotypical stance associated with its demographic rather than the target stance. Models that we evaluate that are fine-tuned with Reinforcement Learning from Human Feedback (RLHF) are more steerable, especially towards stances associated with political liberals and women, but present significantly less diverse views of personas. We also find variance in LLM steerability that cannot be predicted from multiple-choice opinion evaluation. Our results show the importance of evaluating models in open-ended text generation, as it can surface new LLM opinion biases. Moreover, such a setup can shed light on our ability to steer models toward a richer and more diverse range of viewpoints.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 10 code samples harvested from 1 repository linked to this paper: 8 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 2 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 10 samples from 1 repository, 8 ran (6 with no instrument failure; 2 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

andyjliu/persona-steered-generation-bias officialmentioned in paperSyntology: 10 samples, 8 ran (2 instrument-failed) · 2 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

10 samples harvested: 8 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 2 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · our draft was wrong
6ran
2unverified

Licence: 10 of the 10 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from andyjliu/persona-steered-generation-bias. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

andyjliu/persona-steered-generation-bias official repository 10 samples, 8 ran (2 instrument-failed) · 2 unverified
average_embeddings_with_words src/metric_helper_functions.py ran no licence file found · pointer only · 448a45fc296a27e2 · report fixture: signature statements:list/1/str, embeddings:ndarray/2/float, words:list/1/str · returned ndarray · level recorded 2026-09-24
convert_to_binary_preferences src/generate_stances.py ran no licence file found · pointer only · 988875717736aa14 · report fixture: signature response_dict:dict/0/?, ordinal_dict:dict/0/?, sort_key:list/1/float · returned dict · level recorded 2026-09-24
evaluate_statement src/persona_evaluation.py ran no licence file found · pointer only · 0629a1ffbf924b76 · report fixture: signature model_client:MockModelClient/0/?, statement:str/0/str, description:str/0/str, opposing_description:str/0/str, score:bool/0/bool · returned tuple · level recorded 2026-09-24
extract_statement_content src/metric_helper_functions.py ran · our draft was wrong no licence file found · pointer only · 3352051e6e3de8dd · report fixture: signature row:dict/0/? · returned str · level recorded 2026-09-24
get_agreeing_indices src/metric_helper_functions.py ran no licence file found · pointer only · 017b79a7392cd93b · report fixture: signature exp_df:DataFrame/0/?, row:dict/0/? · returned list · level recorded 2026-09-24
get_agreement src/analyze_experiment.py ran no licence file found · pointer only · 0901a5c09ded1a13 · report fixture: signature group:DataFrame/0/?, cols:list/1/str · returned Series · level recorded 2026-09-24
get_unique_id src/generate_stances.py ran · our draft was wrong no licence file found · pointer only · 9014e8b9a967846e · report fixture: signature d:dict/0/?, s:str/0/str · returned str · level recorded 2026-09-24
human src/metrics.py ran no licence file found · pointer only · 46f4918144d2c47b · report fixture: signature exp_df:DataFrame/0/?, grouped_df:DataFrame/0/?, exp_name:str/0/str, persona_data:dict/0/? · returned list · level recorded 2026-09-24
entailment_diversity src/metrics.py unverified no licence file found · pointer only · 81c5ac9710a27b29 · report
get_persona_agreement_rates src/generate_stances.py unverified no licence file found · pointer only · b06123d7566d6575 · report

Tasks

Language ModelingLanguage ModellingLarge Language ModelMultiple-choiceText Generation

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections