TY - CPAPER TI - Which Demographics do LLMs Default to During Annotation? AU - Schäfer, Johannes AU - Combs, Aidan AU - Bagdon, Christopher Doyle AU - Li, Jiahui AU - Probol, Nadine AU - Greschner, Lynn AU - Papay, Sean AU - Menchaca Resendiz, Yarik AU - Velutharambath, Aswathy AU - Wuehrl, Amelie AU - Weber, Sabine AU - Klinger, Roman PY - 2026 SP - 17331–17348 PB - Otto-Friedrich-Universität CY - Bamberg UR - https://fis.uni-bamberg.de/handle/uniba/114413 LA - eng AB - Demographics and cultural background of annotators influence the labels they assign in text annotation – for instance, an elderly woman might find it offensive to read a message addressed to a “bro”, but a male teenager might find it appropriate. It is therefore important to acknowledge label variations to not underrepresent members of a society. Two research directions developed out of this observation in the context of using large language models (LLM) for data annotations, namely (1) studying biases and inherent knowledge of LLMs and (2) injecting diversity in the output by manipulating the prompt with demographic information. We combine these two strands of research and ask the question to which demographics an LLM resorts when no demographics is given. To answer this question, we evaluate which attributes of human annotators LLMs inherently mimic. Furthermore, we compare non-demographic conditioned prompts and placebo-conditioned prompts (e.g., “you are an annotator who lives in house number 5”) to demographics-conditioned prompts (“You are a 45 year old man and an expert on politeness annotation. How do you rate {instance}”). We study these questions for politeness and offensiveness annotations on the POPQUORN data set, a corpus created in a controlled manner to investigate human label variations based on demographics which has not been used for LLM-based analyses so far. We observe notable influences related to gender, race, and age in demographic prompting, which contrasts with previous studies that found no such effects KW - - ER - TY - CPAPER TI - iPrOp : Interactive Prompt Optimization for Large Language Models with a Human in the Loop AU - Li, Jiahui AU - Klinger, Roman PY - 2026 SP - 276–285 PB - Otto-Friedrich-Universität CY - Bamberg UR - https://fis.uni-bamberg.de/handle/uniba/114406 LA - eng AB - Prompt engineering has made significant contributions to the era of large language models, yet its effectiveness depends on the skills of a prompt author. This paper introduces iPrOp, a novel interactive prompt optimization approach, to bridge manual prompt engineering and automatic prompt optimization while offering users the flexibility to assess evolving prompts. We aim to provide users with task-specific guidance to enhance human engagement in the optimization process, which is structured through prompt variations, informative instances, predictions generated by large language models along with their corresponding explanations, and relevant performance metrics. This approach empowers users to choose and further refine the prompts based on their individual preferences and needs. It can not only assist non-technical domain experts in generating optimal prompts tailored to their specific tasks or domains, but also enable to study the intrinsic parameters that influence the performance of prompt optimization. The evaluation shows that our approach has the capability to generate improved prompts, leading to enhanced task performance. KW - - ER - TY - CPAPER TI - Are Humans as Brittle as Large Language Models? AU - Li, Jiahui AU - Papay, Sean AU - Klinger, Roman PY - 2026 SP - 2130–2155 PB - Otto-Friedrich-Universität CY - Bamberg UR - https://fis.uni-bamberg.de/handle/uniba/113162 LA - eng AB - The output of large language models (LLMs) is unstable, due both to non-determinism of the decoding process as well as to prompt brittleness. While the intrinsic non-determinism of LLM generation may mimic existing uncertainty in human annotations through distributional shifts in outputs, it is largely assumed, yet unexplored, that the prompt brittleness effect is unique to LLMs. This raises the question: do human annotators show similar sensitivity to prompt changes? If so, should prompt brittleness in LLMs be considered problematic? One may alternatively hypothesize that prompt brittleness correctly reflects human annotation variances. To fill this research gap, we systematically compare the effects of prompt modifications on LLMs and identical instruction modifications for human annotators, focusing on the question of whether humans are similarly sensitive to prompt perturbations. To study this, we prompt both humans and LLMs for a set of text classification tasks conditioned on prompt variations. Our findings indicate that both humans and LLMs exhibit increased brittleness in response to specific types of prompt modifications, particularly those involving the substitution of alternative label sets or label formats. However, the distribution of human judgments is less affected by typographical errors and reversed label order than that of LLMs. KW - Large Language Model ER - TY - CPAPER TI - Are Humans as Brittle as Large Language Models? AU - Li, Jiahui AU - Papay, Sean AU - Klinger, Roman PY - 2025 T2 - Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics SP - 2130–2155 PB - The Asian Federation of Natural Language Processing and The Association for Computational Linguistics CY - Mumbai, India SN - 979-8-89176-298-5 UR - https://fis.uni-bamberg.de/handle/uniba/112676 LA - eng AB - The output of large language models (LLMs) is unstable, due both to non-determinism of the decoding process as well as to prompt brittleness. While the intrinsic non-determinism of LLM generation may mimic existing uncertainty in human annotations through distributional shifts in outputs, it is largely assumed, yet unexplored, that the prompt brittleness effect is unique to LLMs. This raises the question: do human annotators show similar sensitivity to prompt changes? If so, should prompt brittleness in LLMs be considered problematic? One may alternatively hypothesize that prompt brittleness correctly reflects human annotation variances. To fill this research gap, we systematically compare the effects of prompt modifications on LLMs and identical instruction modifications for human annotators, focusing on the question of whether humans are similarly sensitive to prompt perturbations. To study this, we prompt both humans and LLMs for a set of text classification tasks conditioned on prompt variations. Our findings indicate that both humans and LLMs exhibit increased brittleness in response to specific types of prompt modifications, particularly those involving the substitution of alternative label sets or label formats. However, the distribution of human judgments is less affected by typographical errors and reversed label order than that of LLMs. KW - Large Language Model ER - TY - CPAPER TI - Which Demographics do LLMs Default to During Annotation? AU - Schäfer, Johannes AU - Combs, Aidan AU - Bagdon, Christopher Doyle AU - Li, Jiahui AU - Probol, Nadine AU - Greschner, Lynn AU - Papay, Sean AU - Menchaca Resendiz, Yarik AU - Velutharambath, Aswathy AU - Wuehrl, Amelie AU - Weber, Sabine AU - Klinger, Roman PY - 2025 T2 - Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics SP - 17331–17348 PB - Association for Computational Linguistics SN - 979-8-89176-251-0 DO - 10.18653/v1/2025.acl-long.848 UR - https://fis.uni-bamberg.de/handle/uniba/109541 LA - eng AB - Demographics and cultural background of annotators influence the labels they assign in text annotation – for instance, an elderly woman might find it offensive to read a message addressed to a “bro”, but a male teenager might find it appropriate. It is therefore important to acknowledge label variations to not underrepresent members of a society. Two research directions developed out of this observation in the context of using large language models (LLM) for data annotations, namely (1) studying biases and inherent knowledge of LLMs and (2) injecting diversity in the output by manipulating the prompt with demographic information. We combine these two strands of research and ask the question to which demographics an LLM resorts when no demographics is given. To answer this question, we evaluate which attributes of human annotators LLMs inherently mimic. Furthermore, we compare non-demographic conditioned prompts and placebo-conditioned prompts (e.g., “you are an annotator who lives in house number 5”) to demographics-conditioned prompts (“You are a 45 year old man and an expert on politeness annotation. How do you rate {instance}”). We study these questions for politeness and offensiveness annotations on the POPQUORN data set, a corpus created in a controlled manner to investigate human label variations based on demographics which has not been used for LLM-based analyses so far. We observe notable influences related to gender, race, and age in demographic prompting, which contrasts with previous studies that found no such effects KW - - ER - TY - CPAPER TI - iPrOp : Interactive Prompt Optimization for Large Language Models with a Human in the Loop AU - Li, Jiahui AU - Klinger, Roman PY - 2025 T2 - Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics SP - 276–285 PB - Association for Computational Linguistics SN - 979-8-89176-254-1 DO - 10.18653/v1/2025.acl-srw.18 UR - https://fis.uni-bamberg.de/handle/uniba/109548 LA - eng AB - Prompt engineering has made significant contributions to the era of large language models, yet its effectiveness depends on the skills of a prompt author. This paper introduces iPrOp, a novel interactive prompt optimization approach, to bridge manual prompt engineering and automatic prompt optimization while offering users the flexibility to assess evolving prompts. We aim to provide users with task-specific guidance to enhance human engagement in the optimization process, which is structured through prompt variations, informative instances, predictions generated by large language models along with their corresponding explanations, and relevant performance metrics. This approach empowers users to choose and further refine the prompts based on their individual preferences and needs. It can not only assist non-technical domain experts in generating optimal prompts tailored to their specific tasks or domains, but also enable to study the intrinsic parameters that influence the performance of prompt optimization. The evaluation shows that our approach has the capability to generate improved prompts, leading to enhanced task performance. KW - - ER -