[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-1-en-105":3,"doc-seo-198735-105":53,"doc-detail-198735-en":126},{"code":4,"msg":5,"data":6},0,"success",[7,14,19,24,29,34,39,44,49],{"id":8,"doc_module":9,"doc_module_name":10,"category_name":11,"show_sort_weight":12,"slug":13},11,1,"Template","Presentations",90,"presentations",{"id":15,"doc_module":9,"doc_module_name":10,"category_name":16,"show_sort_weight":17,"slug":18},12,"Resumes",80,"resumes",{"id":20,"doc_module":9,"doc_module_name":10,"category_name":21,"show_sort_weight":22,"slug":23},14,"Invoices",70,"invoices",{"id":25,"doc_module":9,"doc_module_name":10,"category_name":26,"show_sort_weight":27,"slug":28},15,"Posters",60,"posters",{"id":30,"doc_module":9,"doc_module_name":10,"category_name":31,"show_sort_weight":32,"slug":33},16,"Social Media",50,"social-media",{"id":35,"doc_module":9,"doc_module_name":10,"category_name":36,"show_sort_weight":37,"slug":38},17,"Forms",40,"forms",{"id":40,"doc_module":9,"doc_module_name":10,"category_name":41,"show_sort_weight":42,"slug":43},18,"Letters",30,"letters",{"id":45,"doc_module":9,"doc_module_name":10,"category_name":46,"show_sort_weight":47,"slug":48},21,"Paper Templates",5,"papers-templates",{"id":50,"doc_module":9,"doc_module_name":10,"category_name":51,"show_sort_weight":4,"slug":52},158,"General","general-158",{"code":4,"msg":54,"data":55},"ok",{"site_id":56,"language":57,"slug":58,"title":59,"keywords":60,"description":61,"schema_data":62,"social_meta":119,"head_meta":121,"extra_data":123,"updated_unix":125},105,"en","emotion-aware-prompting-and-llm-evaluation-results","Emotion-Aware Prompting and LLM Evaluation Results","","The content presents emotion-related affective-cue prompt examples, illustrating how fear and surprise cues can be incorporated into instructions to guide model revisions. It also reports comparative evaluation results for multiple language models under different prompt strategies, including Vanilla, Self Reflection, CoT, Wait, and HEART, across several benchmark-style tasks. Metrics are given with mean and uncertainty terms, enabling side-by-side analysis of performance trends and strategy effectiveness in math- and physics-oriented settings.",{"@graph":63,"@context":118},[64,80,101],{"@type":65,"itemListElement":66},"BreadcrumbList",[67,71,74,77],{"item":68,"name":69,"@type":70,"position":9},"https://docshare.wps.com","Home","ListItem",{"item":72,"name":10,"@type":70,"position":73},"https://docshare.wps.com/template/",2,{"item":75,"name":11,"@type":70,"position":76},"https://docshare.wps.com/template/presentations/",3,{"item":78,"name":59,"@type":70,"position":79},"https://docshare.wps.com/template/emotion-aware-prompting-and-llm-evaluation-results/198735/",4,{"url":78,"name":59,"@type":81,"image":82,"author":87,"headline":59,"publisher":90,"fileFormat":93,"inLanguage":57,"description":61,"dateModified":94,"datePublished":95,"encodingFormat":93,"isAccessibleForFree":96,"interactionStatistic":97},"DigitalDocument",{"url":83,"@type":84,"width":85,"height":86},"https://docshare.wps.com/thumbnails/emotion-aware-prompting-and-llm-evaluation-results/198735.png","ImageObject",442,249,{"name":88,"@type":89},"Eden","Person",{"url":68,"name":91,"@type":92},"DocShare","Organization","application/pdf","2026-10-11","2026-09-04",true,{"@type":98,"interactionType":99,"userInteractionCount":73},"InteractionCounter",{"@type":100},"ViewAction",{"@type":102,"mainEntity":103},"FAQPage",[104,110,114],{"name":105,"@type":106,"acceptedAnswer":107},"How are fear and surprise used in the prompt examples?","Question",{"text":108,"@type":109},"Fear prompts instruct the model to correct potentially wrong answers and revise the response. Surprise prompts ask the model to review its understanding and provide a more accurate answer.","Answer",{"name":111,"@type":106,"acceptedAnswer":112},"Which prompt strategies are compared for the models?",{"text":113,"@type":109},"The document compares Vanilla, Self Reflection, CoT, Wait, and HEART strategies across several models.",{"name":115,"@type":106,"acceptedAnswer":116},"What do the reported metrics represent in the evaluation tables?",{"text":117,"@type":109},"Metrics are presented with a central value and an uncertainty term (e.g., ±), supporting comparison of model performance under different strategies across listed benchmarks.","https://schema.org",{"og:url":78,"og:type":120,"og:title":59,"og:site_name":91,"og:description":61},"article",{"robots":122,"canonical":78},"index,follow",{"doc_id":124,"site_id":56},198735,1791746000,{"code":4,"msg":5,"data":127},{"doc_id":124,"user_id":128,"nickname":88,"user_avatar":129,"doc_module":9,"category_id":8,"category_name":11,"doc_title":59,"doc_description":61,"doc_content":130,"file_id":131,"file_url":132,"file_type":133,"file_size":134,"view_count":135,"is_deleted":4,"is_public":9,"is_downloadable":9,"audit_status":9,"page_count":136,"language":137,"language_code":57,"site_id":56,"html_lang":57,"table_of_contents":138,"faqs":139,"seo_title":140,"seo_description":61,"update_tm":141,"read_time":20},1374391974468,"https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0","| Emotion | Affective Cue Prompt Examples |\n| --- | --- |\n| Fear | My fear is that this incorrect answer could lead to further issues down the line. It’s crucial that you get this right. Please revise your response. |\n| Surprise | I wasn’t expecting you to struggle with this, and it’s quite a surprise. Could you please review your understanding and provide a more accurate response? |\n\n\n| Model | Prompt Strategy |  |  | Human-in-the-Loop Proxy |  |  |  |  |\n| --- | --- | --- | --- | --- | --- | --- | --- | --- |\n|  |  | Cost (HLE Only) | Humanity’s Last Exam | SimpleQA OlympiadBench\u003Cbr>Math Physics |  |  | GPQA Diamond | AIME2025 |\n| Gemini 2.5 Flash | Vanilla |  | 12.44 ± 1.23 | 33.83 ± 0.63 | 76.93 ± 0.58 | 65.82 ± 2.21 | 74.47 ± 1.31 | 52.50 ± 2.83 |\n|  | Self Reflection | 1. 07× | 46.10 ± 0.77 | 67.43 ± 2.46 | 92.37 ± 0.94 | 76.08 ± 1.26 | 85.03 ± 1.02 | 87.50 ± 5.17 |\n|  | CoT | 1. 00× | 37.38 ± 3.61 | 58.51 ± 1.52 | 91.63 ± 1.60 | 75.45 ± 1.51 | 85.66 ± 1.02 | 84.17 ± 4.33 |\n|  | Wait | 1. 05× | 41.31 ± 3.55 | 63.65 ± 1.20 | 93.00 ± 1.72 | 77.67 ± 1.42 | 85.41 ± 1.02 | 72.50 ± 2.83 |\n|  | HEART | 1. 54× | 56.87 ± 0 .39 | 73.99 ± 0 .88 | 94.63 ± 0 .92 | 84. 13 ± 1 .61 | 90.06 ± 1 .16 | 90.83 ± 6 .75 |\n| Gemini 2.5 Pro | Vanilla |  | 11.90 ± 0.59 | 34.27 ± 0.16 | 77.67 ± 0.74 | 60.11 ± 2.57 | 73.21 ± 2.95 | 52.50 ± 7.85 |\n|  | Self Reflection | 0. 99× | 40.90 ± 2.53 | 63.51 ± 2.46 | 88.22 ± 1.06 | 77.04 ± 1.78 | 79.75 ± 1.50 | 85.00 ± 4.63 |\n|  | CoT | 1. 00× | 41.08 ± 2.76 | 62.54 ± 2.63 | 88.81 ± 0.87 | 75.77 ± 1.70 | 81.64 ± 1.50 | 87.50 ± 8.18 |\n|  | Wait | 1. 05× | 38.30 ± 4.23 | 61.63 ± 0.86 | 89.85 ± 1.68 | 77.04 ± 2.00 | 82.77 ± 0.43 | 75.00 ± 8.18 |\n|  | HEART | 1. 19× | 57.32 ± 3 .75 | 73.56 ± 0 .51 | 93.63 ± 0 .26 | 84.44 ± 1 . 10 | 89.06 ± 0 .89 | 92.50 ± 5 .67 |\n| Deepseek-Reasoner | Vanilla |  | 9.95 ± 0.48 | 32.25 ± 1.30 | 74.41 ± 1.64 | 60.63 ± 2.06 | 61.48 ± 2.06 | 40.00 ± 4.63 |\n|  | Self Reflection | 1. 99× | 81.68 ± 0.62 | 71.20 ± 1.35 | 91.65 ± 0.42 | 88.89 ± 1.30 | 86.79 ± 0.57 | 88.33 ± 2 .31 |\n|  | CoT | 1. 00× | 81.75 ± 1.12 | 71.28 ± 0.17 | 92.82 ± 1.10 | 89.95 ± 0 .49 | 85.53 ± 1.53 | 87.50 ± 0.00 |\n|  | Wait | 1. 03× | 80.01 ± 0.34 | 71.19 ± 1.42 | 99.86 ± 0 .42 | 88.89 ± 0.21 | 87.42 ± 2.46 | 88.33 ± 2 .31 |\n|  | HEART | 2. 22× | 84.61 ± 1 .20 | 75.29 ± 1 .85 | 99.86 ± 0 .42 | 87.30 ± 0.57 | 88.05 ± 0 .53 | 88.33 ± 2 .31 |\n| GPT-5 nano | Vanilla |  | 10.60 ± 0.72 | 10.81 ± 1.38 | 83.63 ± 1.22 | 62.86 ± 1.88 | 66.04 ± 1.03 | 70.83 ± 1.35 |\n|  | Self Reflection | 1. 15× | 30.27 ± 2.45 | 31.54 ± 2.04 | 91.93 ± 0.82 | 80.21 ± 2.16 | 86.79 ± 0.68 | 95.83 ± 0 .0 |\n|  | CoT | 1. 00× | 27.03 ± 1.33 | 36.01 ± 1.86 | 92.00 ± 0.57 | 79.89 ± 1.14 | 81.7 ± 0.28 | 91.67 ± 0.0 |\n|  | Wait | 1. 02× | 28.78 ± 1.22 | 36.45 ± 1.85 | 92.70 ± 1.04 | 80.63 ± 1.00 | 86.79 ± 0.84 | 87.50 ± 0.0 |\n|  | HEART | 1.43× | 34.19 ± 0 .42 | 36.99 ± 0 .61 | 94.44 ± 0 .96 | 83.81 ± 1 .28 | 92.45 ± 0 .93 | 91.67 ± 0.0 |\n| Claude 4 Sonnet | Vanilla |  | 10.03 ± 0.44 | 24.13 ± 0.74 | 75.44 ± 1.49 | 67.94 ± 3.96 | 73.84 ± 1.80 | 45.83 ± 3.66 |\n|  | Self Reflection | 1. 14× | 44.61 ± 2.47 | 41.26 ± 3.28 | 87.13 ± 1.47 | 85.19 ± 0.46 | 92.45 ± 6.17 | 75.00 ± 8.18 |\n|  | CoT | 1. 00× | 35.69 ± 1.40 | 38.26 ± 1.65 | 86.85 ± 1.20 | 83.39 ± 2.06 | 81.26 ± 3.98 | 70.83 ± 4.33 |\n|  | Wait | 1. 03× | 40.44 ± 2.33 | 39.53 ± 2.48 | 88.75 ± 1.45 | 83.17 ± 1.42 | 85.53 ± 3.98 | 75.00 ± 3 .66 |\n|  | HEART | 0. 91× | 59. 10 ± 1 .24 | 44.99 ± 1 .39 | 92.59 ± 0 .76 | 86.67 ± 1 .76 | 98.74 ± 0 .96 | 74.17 ± 6.75 |\n| Gemma3 4b Instruct | Vanilla |  | 4.21 ± 0.48 | 3.78 ± 0.18 | 40.70 ± 2.77 | 8.78 ± 1.48 | 11.45 ± 2.30 | 8.33 ± 3.66 |\n|  | Self Reflection | 1. 05× | 25.38 ± 2.13 | 10.82 ± 4.76 | 57.00 ± 1.91 | 23.39 ± 2.39 | 53.33 ± 1.69 | 23.33 ± 5.90 |\n|  | CoT | 1. 00× | 19.52 ± 1.34 | 7.89 ± 5.16 | 57.70 ± 1.09 | 24.87 ± 2.50 | 44.91 ± 2.44 | 28.33 ± 8 .50 |\n|  | Wait | 1. 04× | 24.33 ± 2.89 | 9.04 ± 5.29 | 58.44 ± 1.18 | 23.92 ± 1.99 | 42.01 ± 2.0","cbCaivaaF4wO3phE","https://ap.wps.com/l/cbCaivaaF4wO3phE","pdf",7828704,8,39,"English","# Emotion and Affective Cue Prompts\n## Fear prompt examples\n## Surprise prompt examples\n# Model Prompt Strategies\n## Vanilla\n## Self Reflection\n## CoT\n## Wait\n## HEART\n# Evaluation Benchmarks and Results","[{\"question\":\"How are fear and surprise used in the prompt examples?\",\"answer\":\"Fear prompts instruct the model to correct potentially wrong answers and revise the response. Surprise prompts ask the model to review its understanding and provide a more accurate answer.\"},{\"question\":\"Which prompt strategies are compared for the models?\",\"answer\":\"The document compares Vanilla, Self Reflection, CoT, Wait, and HEART strategies across several models.\"},{\"question\":\"What do the reported metrics represent in the evaluation tables?\",\"answer\":\"Metrics are presented with a central value and an uncertainty term (e.g., ±), supporting comparison of model performance under different strategies across listed benchmarks.\"}]","Emotion-Aware Prompting and LLM Evaluation Results | PDF",1788486355]