Pith. sign in

REVIEW 5 major objections 6 minor 60 references

VicSim: Enhancing Victim Simulation with Emotional and Linguistic Fidelity

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read VicSim claims that a small fine-tuned language model generates victim reports that human raters find as human-like as real victims' messages and more human-like than GPT-4.

desk verdict A genuinely useful domain application and some solid empirical observations, but the headline 'outperforms GPT-4' rests on a human eval that lacks direct comparisons, sample sizes, and at least one internally consistent statistic. read the letter →

arxiv 2501.03139 v1 pith:GWZE52Y3 submitted 2025-01-06 cs.CL cs.HC

Yerong Li , Yiren Liu , Yun Huang This is my paper · ORCID
classification cs.CLcs.HC
keywords victimsimulationscenario-basedtrainingusersafetyincidentreportinglargelanguagemodelsadversarialgrammarstylehuman-likeness
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

VicSim is a victim simulator for text-based safety-incident reporting. The paper tries to show that a small, locally fine-tuned language model can produce victim utterances that human readers judge to be as human-like as real victims' messages, and more human-like than the much larger commercial model used as a baseline. That claim matters because dispatcher training now depends on human role-players; a convincing automated victim could make scenario-based training cheaper, more repeatable, and easier to scale across institutions. To get there, the paper combines two training signals that are unusual for language generation: it teaches the model to make grammar mistakes like real users, and it rewards emotional expression that tracks the arc of a real incident report. The paper also claims the simulator is more faithful to scenario facts than the commercial baseline, with fewer hallucinations.

What carries the argument

The load-bearing mechanism is a generator–discriminator loop, a GAN-style arrangement in which one network tries to produce text that fools a second network. The generator is a 7-billion-parameter chat model fine-tuned on real text-based incident-reporting dialogues; the discriminator is a smaller instruction-tuned classifier that is first trained on emotion and grammar labels, then trained adversarially to tell real victim utterances from generated ones. Because punctuation errors dominate human messages, the discriminator learns to treat missing periods and similar slips as evidence of humanness, and the generator is updated to reproduce those slips. A second ingredient is key-information prompting: an information-extraction toolkit pulls named entities and other facts from each scenario, and those keywords are appended to the prompt so the generator reproduces them instead of inventing details. The discriminator's grammar and emotion sensitivity carries the human-likeness result, while the keyword prompting carries the faithfulness result.

What would settle it

A preregistered human evaluation with a fixed sample size and an equivalence test would settle the claim: if raters, given enough incidents and enough raters, can identify VicSim responses as computer-generated at a rate clearly above chance in a forced-choice test, then the central human-likeness claim is false. The same test should also check whether the result survives when punctuation errors are normalized out of the responses.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central finding is that a 7-billion-parameter chat model fine-tuned on real incident-report dialogues and trained against an adversarial classifier can produce victim responses that human raters rate no differently from human-written responses, while rating GPT-4's responses as significantly more likely AI-generated. The paper also reports that adding extracted scenario keywords to the prompt raises recall of key facts and reduces hallucination relative to GPT-4, and that the simulated emotional trajectory over the conversation matches the human pattern of negative emotion concentrated early in the report. Grammar style is treated as a first-class signal: adversarial training with a discriminator tuned to spot grammar errors and emotional wording pushes the generator toward the punctuation errors and informal register that dominate real user messages, where more than 88% of unmodified LLM utterances are error-free but fewer than 6% of human utterances are. These three strands—informational faithfulness, emotional dynamics, and language style—are the paper's stated dimensions of human-like victim simulation.

Load-bearing premise

The conclusion that VicSim is human-like depends on a human-rating study in which each rater judged one randomly sampled incident; the paper does not report the number of raters, the number of incidents sampled, inter-rater agreement, or a power analysis, so the non-significant difference between VicSim and human-written responses could come from low statistical power rather than genuine indistinguishability.

Editorial extensions

If this is right

  • Dispatcher-training systems could use VicSim as a stand-in for human role-players in text-based incident-reporting practice, since human raters find its responses as human-like as real victims' reports.
  • Trainees would experience the realistic emotional arc of an incident report—negative affect weighted toward the first part of the conversation—rather than the emotionally flat profile of a generic chatbot.
  • Prompting with extracted scenario keywords should reduce hallucinations, so trainees receive consistent incident details across repeated practice runs.
  • The adversarial loop turns grammatical imperfection into a modeled behavior, so the simulator is deliberately taught to write like a hurried civilian rather than a polished assistant.
  • If these results hold, text-based safety-reporting systems can generate diverse, repeatable training scenarios without collecting new sensitive victim reports.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: if grammar errors and emotional cues are what make synthetic victims believable, then the usual goal of polished, fluent LLM output is the wrong target for user simulation; training objectives for this task should explicitly reward imperfect, affect-laden text.
  • Editorial inference: the same generator–discriminator-plus-keywords recipe could transfer to other personas whose surface style matters—frustrated customers, non-native speakers, elderly users—but the paper only demonstrates it for safety-incident victims, so that transfer is untested.
  • Editorial inference: because the discriminator leans heavily on missing punctuation, VicSim's apparent human-likeness may be partly a surface effect; a targeted test would strip punctuation cues from evaluation and see whether human-likeness ratings persist.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 6 minor

Summary. This paper presents VicSim, a Llama-2-7B-based victim simulator for text-based safety incident reporting. The authors fine-tune the generator with a GAN-style objective using an instruction-tuned Flan-T5 discriminator that is trained to recognize grammar errors and emotional cues, and they augment prompts with CoreNLP key information to reduce hallucination. They evaluate informational faithfulness, emotional dynamics, grammar style, and human-perceived human-likeness against human-written responses and an off-the-shelf GPT-4 baseline, concluding in the abstract that human raters found VicSim more human-like than GPT-4.

Significance. If the headline claim were supported, the paper would be a useful demonstration that a small fine-tuned model can match human style in a constrained domain and that adversarial training with grammar/emotion cues plus key-information prompting improves fidelity. The application area (dispatcher training for text-based incident reporting) is under-explored, and the use of a real anonymized dataset from a deployed system is a strength, as are the qualitative examples and the attempt to evaluate human-likeness directly. However, the current evidence is not sufficient: the human evaluation is under-powered and under-reported, the comparative baseline is not controlled, one key correlation result contradicts its own numbers, and several automatic fidelity results are partly by construction of the training objective. The contribution is therefore suggestive rather than demonstrated.

major comments (5)
  1. [3.3.4 and Figure 6] The human evaluation does not support the abstract's claim that VicSim outperforms GPT-4 in human-likeness. The manuscript reports no sample size (number of raters or incidents), no inter-rater agreement, no power or equivalence analysis, and no direct VicSim-versus-GPT-4 significance test. The only inferential statistic reported is a 'paired t-test' with F=-2.22 for human versus GPT-4, which is not a t statistic, and a null human-VicSim result that cannot be interpreted as evidence of similarity without an equivalence test or confidence interval. This is the central evidence for the headline claim, so the analysis must be fully reported and re-run with appropriate tests.
  2. [4.3 and Figure 9] The sentence 'VicSim has a stronger positive correlation coefficient with humans (r = 0.66, p = 0.03*) than that of GPT-4 (r = 0.88, p < 0.001***)' is internally contradictory: 0.88 is larger than 0.66, so the data as reported show that GPT-4 has the stronger correlation. This directly undermines the grammar-style fidelity claim. The authors should re-check the numbers, clarify what is being correlated, and correct the claim and the figure caption.
  3. [3.1, Eqs. (1)-(2)] The GAN formulation is specified only for continuous outputs; no gradient estimator or training algorithm is provided for the discrete token sampling of the Llama-2 generator. Without specifying how the generator loss in Eq. (1) is backpropagated through discrete text (e.g., policy gradient, Gumbel-softmax, or a distillation approach), the method is not reproducible, and the causal claims about what the adversarial training 'teaches' the generator cannot be verified.
  4. [3.1 and 4.1] The comparative claim against GPT-4 is confounded: VicSim is fine-tuned on 5,427 domain utterances and receives key-information prompting, while GPT-4 appears to be used off-the-shelf. To support the statement that VicSim 'outperforms GPT-4,' the authors need a controlled baseline, such as GPT-4 fine-tuned on the same data, or at minimum GPT-4 with the same key-information prompting and system prompt. As reported, the comparison conflates fine-tuning and prompting with inherent model quality.
  5. [3.1 and 4.2/4.3] The automatic grammar and emotion fidelity results are in part by construction: the discriminator is explicitly instruction-tuned on grammar and emotion cues (Section 3.1), and the generator is optimized to fool that discriminator via Eqs. (1)-(2). Observing that VicSim's punctuation-error rate and emotional expression pattern move closer to humans is therefore an expected optimization outcome rather than independent evidence of human-likeness. These analyses should be reframed as diagnostic checks of the training objective, and the human evaluation must carry the weight of the human-likeness claim.
minor comments (6)
  1. [3.1] The two consecutive paragraphs describing the prompt construction are nearly identical and should be merged into one.
  2. [4.2.2 and Table 3] There are typos: 'sbout' should be 'about', and the Table 3 caption says 'When asked to generated successive responses' instead of 'to generate.'
  3. [Figures 3 and 5] 'Senario' in the Figure 3 caption should be 'Scenario.'
  4. [4.1 and 4.2] Test statistics are labeled F for what appear to be paired t-tests or other comparisons (e.g., F=-2.22, F=-6.94); please use the correct statistic names and report degrees of freedom or effect sizes.
  5. [4.2.2] The phrase 'VicSim, with or without adversarial training, ... the length and count correlate more strongly, with values of 0.44 and 0.57' is ambiguous about which condition corresponds to which value; please label the conditions clearly.
  6. [Template and references] The ACM template still contains placeholder text ('Make sure to enter the correct conference title from your rights confirmation email') and the copyright year 2018; these should be updated before any submission.

Circularity Check

1 steps flagged · score 4.0 of 10

Grammar and emotion fidelity findings partly reduce to the discriminator training objective, but the headline human-likeness claim rests on independent human ratings and is not circular.

  1. fitted input called prediction [Section 3.1 (Eqs. 1-2) and Section 4.3]
    "To improve the text representation of the Flan-T5-based encoder, we instruction-tuned Flan-T5 with emotional prompts and prompts designed to simulate grammar errors. [...] Thus, from Figure 9, we see a tendency in the results of adversarial training where VicSim is encouraged to generate punctuation errors to spoof the discriminator."

    The generator loss in Eq. (1) is -E[log D(G(p))], so the generator is explicitly optimized to reproduce whatever cues the discriminator was fitted to reward. Section 3.1 says that discriminator was instruction-tuned on emotional and grammar-error prompts, and Section 4.3 shows punctuation errors are the cue it exploits. Reporting the resulting punctuation-error behavior as evidence of human-like grammar fidelity is therefore reporting the training objective's own output, not an independent prediction. The finding is not fully circular because the human-likeness claim additionally relies on separate human raters, but the grammar/emotion automatic evaluations are partly self-confirming.

full rationale

No load-bearing self-citation or imported-uniqueness circularity is present: citations [30] and [34] supply dataset and background context, and no prior result is invoked to forbid alternatives. The paper's headline claim that VicSim is more human-like than GPT-4 depends on human ratings, which are external to the training loss and so are not circular, although the reported statistics are under-powered and the inference from two separate tests is a correctness concern rather than a circularity concern. The one genuine circular component is the grammar/emotion fidelity evaluation: the Flan-T5 discriminator is instruction-tuned on grammar and emotion cues, and the generator is trained by Eq. (1) to fool it, so the observed punctuation-error and emotional-word behaviors in Sections 4.2-4.3 are training-objective echoes. Because the central human-likeness claim retains independent content but the grammar/emotion sub-claims partially reduce by construction, the overall circularity score is 4.

Assumptions & free parameters 3 free parameters · 5 assumptions · 1 invented entities

The ledger shows that VicSim depends heavily on learned model weights, unreported training hyperparameters, LLM-generated scenario summaries used as ground truth, and several domain assumptions about what makes text human-like. The central contribution is therefore an engineered system whose behavior is shaped by its own training signal, with limited independent anchoring outside the paper's own human evaluation.

free parameters (3)
  • Fine-tuned model weights of VicSim generator and Flan-T5 discriminator = Not reported in the paper.
    The central system is defined by learned weights from Llama-2 7B and Flan-T5. Without checkpoints or an exact training procedure, the parameter values behind the reported behavior cannot be independently inspected.
  • GAN training hyperparameters = Not reported in the paper.
    Learning rates, batch size, number of adversarial rounds, and any reward scaling for the discrete-text generator are absent, so the trained system cannot be reproduced from the text alone.
  • RoBERTa grammar and emotion classifier configurations = Not reported in the paper.
    The paper reports 94.25% accuracy for the grammar classifier and uses GoEmotions for emotion classification, but does not provide training splits, class balance, or thresholds used for evaluation.
assumptions (5)
  • domain assumption Scenario summaries produced by Llama-2 70B are faithful ground-truth representations of victim reports.
    These summaries are used both for prompt construction and for keyword-overlap faithfulness evaluation (Sections 3.1 and 3.3.1). If the summaries omit or invent details, both the training signal and the faithfulness metric are biased.
  • domain assumption The GAN discriminator's grammar and emotion judgments are valid reward signals for making an LLM more human-like.
    The generator is tuned to fool a Flan-T5 discriminator trained on grammar and emotion cues (Section 3.1), so gains on those cues are partly a response to the reward rather than independent evidence about human realism.
  • domain assumption A single final user utterance rated by crowdworkers is a sufficient measure of human-likeness for dispatcher training.
    The human evaluation presents one last response with dialogue history and collects Likert ratings (Section 3.3.4). No evidence connects these ratings to training outcomes, and the paper acknowledges that no actual training evaluation was conducted (Section 6).
  • domain assumption Matching human grammar error frequencies, especially punctuation errors, makes a simulated victim more human-like.
    The adversarial training emphasizes punctuation and grammar cues (Sections 4.3 and 5.1). The paper does not show that deliberate punctuation errors improve dispatcher performance rather than merely fooling raters.
  • ad hoc to paper The standard GAN objective in Eqs (1)-(2) can be applied to discrete text without a specified gradient estimator.
    The paper states GAN losses but does not describe how token-level outputs from Llama-2 backpropagate or how the generator is updated, so the training mechanism is under-specified.
invented entities (1)
  • Synthetic scenario summaries and filled-in entity names
    purpose: Serve as prompt context and evaluation ground truth for keyword faithfulness, replacing real anonymized victim narratives.
    These summaries are synthesized by Llama-2 70B from masked data (Section 3.2.2) and used both to train and to score the model (Figure 5). There is no external validation that they reflect what real victims would say, so they do not provide independent evidence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VicSim: Enhancing Victim Simulation with Emotional and Linguistic Fidelity." pith.science (2026). https://pith.science/paper/GWZE52Y3

@misc{pith2026250103139,
  author       = {Pith},
  title        = {Pith review of: VicSim: Enhancing Victim Simulation with Emotional and Linguistic Fidelity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GWZE52Y3}},
  note         = {Machine review of arXiv:2501.03139}
}
read the original abstract

Scenario-based training has been widely adopted in many public service sectors. Recent advancements in Large Language Models (LLMs) have shown promise in simulating diverse personas to create these training scenarios. However, little is known about how LLMs can be developed to simulate victims for scenario-based training purposes. In this paper, we introduce VicSim (victim simulator), a novel model that addresses three key dimensions of user simulation: informational faithfulness, emotional dynamics, and language style (e.g., grammar usage). We pioneer the integration of scenario-based victim modeling with GAN-based training workflow and key-information-based prompting, aiming to enhance the realism of simulated victims. Our adversarial training approach teaches the discriminator to recognize grammar and emotional cues as reliable indicators of synthetic content. According to evaluations by human raters, the VicSim model outperforms GPT-4 in terms of human-likeness.

Figures

Figures reproduced from arXiv: 2501.03139 by the authors.

Figure 1
Figure 1. GAN training workflow : we used Flan-T5 based discriminator and Llama-2 chat-based simulated user as the generator [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the prompt construction process for simulated user dialogue generation: we concatenate the system guidance, [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Senario-based prompt argumented by the key information extracted from the CoreNLP toolkit [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Users’ utterances at different conversation stages on the evaluation set [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Keyword-based assessment on hallucination, here [John Smith], [Cortright] are simulated key information we filled in for [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Distribution of Ratings from Human Raters; paired t-test indicated a significant difference between the responses from human [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Emotional Trajectories of Human and Simulated Users Throughout Dialogues: with or without adversarial training, local LLM [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Difference in distribution of emotional keywords over human/LLM responses of different lengths [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Type of the errors made by humans and LLMs. Only 4.12% of human utterances make no grammar mistakes, while more than [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 47 canonical work pages

  1. [1]

    Krisztian Balog and ChengXiang Zhai. 2023. User Simulation for Evaluating Information Access Systems. In Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retrieval in the Asia Pacific Region (, Beijing, China,) (SIGIR-AP ’23). Association for Computing Machinery, New York, NY, USA, 302–305. https://do...

    arXiv 2023

  2. [2]

    Markus Bayer, Marc-André Kaufhold, and Christian Reuter. 2022. A survey on data augmentation for text classification. Comput. Surveys 55, 7 (2022), 1–39

    2022

  3. [3]

    Christian Becker, Stefan Kopp, and Ipke Wachsmuth. 2004. Simulating the emotion dynamics of a multimodal conversational agent. In tutorial and research workshop on affective dialogue systems . Springer, 154–165

    work page 2004

  4. [4]

    Patrik Berger and Michal Kompan. 2019. User modeling for churn prediction in E-commerce. IEEE Intelligent Systems 34, 2 (2019), 44–52

    work page 2019

  5. [5]

    Alex Beutel. 2016. User behavior modeling with large-scale graph analysis. Computer Science Department, Carnegie Mellon University (2016)

    work page 2016

  6. [6]

    Lachlan Birdsey, Claudia Szabo, and Yong Meng Teo. 2015. Twitter knows: understanding the emergence of topics in social networks. In2015 Winter Simulation Conference (WSC). IEEE, 4009–4020

    work page 2015

  7. [7]

    Anne De la Croix and John Skelton. 2013. The simulation game: an analysis of interactions between students and simulated patients. Medical education 47, 1 (2013), 49–58

    work page 2013

  8. [8]

    Gustavo H De Rosa and João P Papa. 2021. A survey on text generation using generative adversarial networks. Pattern Recognition 119 (2021), 108098

    work page 2021

Show all 60 references
  1. [9]

    Yashar Deldjoo, Dietmar Jannach, Alejandro Bellogin, Alessandro Difonzo, and Dario Zanzonelli. 2024. Fairness in recommender systems: research landscape and future directions. User Modeling and User-Adapted Interaction 34, 1 (2024), 59–108

    2024

  2. [10]

    Orianna Demasi, Yu Li, and Zhou Yu. 2020. A multi-persona chatbot for hotline counselor training. In Findings of the Association for Computational Linguistics: EMNLP 2020. 3623–3636

    2020

  3. [11]

    David Donahue and Anna Rumshisky. 2018. Adversarial text generation without reinforcement learning. arXiv preprint arXiv:1810.06640 (2018)

    2018 arXiv

  4. [12]

    Jasper Feine, Ulrich Gnewuch, Stefan Morana, and Alexander Maedche. 2019. A taxonomy of social cues for conversational agents. International Journal of Human-Computer Studies 132 (2019), 138–161

    2019

  5. [13]

    Heidi Kevoe Feldman. 2021. Calming emotional 911 callers: Using redirection as a patient-focused directive in emergency medical calls. Language & Communication 81 (2021), 81–92

    2021

  6. [14]

    Shangsheng Gao, Li Gao, Qi Li, and Jianjun Xu. 2023. Application of large language model in intelligent Q&A of digital government. In Proceedings of the 2023 2nd International Conference on Networks, Communications and Information Technology . 24–27

    2023

  7. [15]

    Maíra Gatti, Paulo Cavalin, Samuel Barbosa Neto, Claudio Pinhanez, Cícero dos Santos, Daniel Gribel, and Ana Paula Appel. 2014. Large-scale multi-agent-based modeling and simulation of microblogging-based online social network. In Multi-Agent-Based Simulation XIV: Internationa...

    2014

  8. [16]

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative Adversarial Nets. In Advances in Neural Information Processing Systems , Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, and K.Q...

    2014

  9. [17]

    Richard Gray. 2011. Lies, liars, and lie detection. Fed. Probation 75 (2011), 31. Manuscript submitted to ACM 20 Yerong Li, Yiren Liu, and Yun Huang

    2011

  10. [18]

    Izzeddin Gür, Dilek Hakkani-Tür, Gokhan Tür, and Pararth Shah. 2018. User modeling for task oriented dialogues. In 2018 IEEE Spoken Language Technology Workshop (SLT). IEEE, 900–906

    2018

  11. [19]

    Xiaomeng Hu, Pin-Yu Chen, and Tsung-Yi Ho. 2023. Radar: Robust ai-text detection via adversarial learning. Advances in Neural Information Processing Systems 36 (2023), 15077–15095

    2023

  12. [20]

    Hutto and Eric Gilbert

    C. Hutto and Eric Gilbert. 2014. VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text. Proceedings of the International AAAI Conference on Web and Social Media 8, 1 (May 2014), 216–225. https://doi.org/10.1609/icwsm.v8i1.14550

    2014 doi

  13. [21]

    Hang Jiang, Xiajie Zhang, Xubo Cao, Jad Kabbara, and Deb Roy. 2023. Personallm: Investigating the ability of gpt-3.5 to express personality traits and gender differences. arXiv preprint arXiv:2305.02547 (2023)

    2023 arXiv

  14. [22]

    Eunkyung Jo, Daniel A Epstein, Hyunhoon Jung, and Young-Ho Kim. 2023. Understanding the benefits and challenges of deploying conversational AI leveraging large language models for public health intervention. In Proceedings of the 2023 CHI Conference on Human Factors in Computi...

    2023

  15. [23]

    Kaggle. 2021. Grammar Correction Dataset. https://www.kaggle.com/datasets/satishgunjal/grammar-correction

    2021

  16. [24]

    Simon Kolmanič, Nikola Guid, and Andrej Nerat. 2013. SIN: Multimedia-based teaching tool for computer-supported fire-fighter training. Fire safety journal 61 (2013), 26–35

    2013

  17. [25]

    Tzu-Sheng Kuo, Hong Shen, Jisoo Geum, Nev Jones, Jason I Hong, Haiyi Zhu, and Kenneth Holstein. 2023. Understanding frontline workers’ and unhoused individuals’ perspectives on ai used in homeless services. In Proceedings of the 2023 CHI Conference on Human Factors in Computin...

    2023

  18. [26]

    Jason Li and Abdolreza Abhari. 2017. Generating stochastic data to simulate a twitter user. In Proceedings of the 20th Communications & Networking Symposium. 1–11

    2017

  19. [27]

    Shuokai Li, Ruobing Xie, Yongchun Zhu, Xiang Ao, Fuzhen Zhuang, and Qing He. 2022. User-centric conversational recommendation with multi-aspect user modeling. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval . 223–233

    2022

  20. [28]

    Xiujun Li, Zachary C Lipton, Bhuwan Dhingra, Lihong Li, Jianfeng Gao, and Yun-Nung Chen. 2016. A User Simulator for Task-Completion Dialogues. arXiv preprint arXiv:1612.05688 (2016)

    2016 arXiv

  21. [29]

    Jionghao Lin, Danielle R Thomas, Feifei Han, Shivang Gupta, Wei Tan, Ngoc Dang Nguyen, and Kenneth R Koedinger. 2023. Using large language models to provide explanatory feedback to human tutors. arXiv preprint arXiv:2306.15498 (2023)

    2023 arXiv

  22. [30]

    Yiren Liu, Ryan Mayfield, and Yun Huang. 2023. Discovering the Hidden Facts of User-Dispatcher Interactions via Text-based Reporting Systems for Community Safety. Proceedings of the ACM on Human-Computer Interaction 7, CSCW1 (2023), 1–31

    2023

  23. [31]

    Kimberly Lomis, Pamela Jeffries, Anthony Palatta, Melanie Sage, Javaid Sheikh, Carl Sheperis, and Alison Whelan. 2021. Artificial intelligence for health professions educators. NAM perspectives 2021 (2021)

    2021

  24. [32]

    Shayne Longpre, Le Hou, Tu Vu, Albert Webson, Hyung Won Chung, Yi Tay, Denny Zhou, Quoc V Le, Barret Zoph, Jason Wei, et al. 2023. The flan collection: Designing data and methods for effective instruction tuning. In International Conference on Machine Learning . PMLR, 22631–22648

    2023

  25. [33]

    Christopher Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven Bethard, and David McClosky. 2014. The Stanford CoreNLP Natural Language Processing Toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations , Ka...

    2014 doi

  26. [34]

    Shufan Ming, Ryan DW Mayfield, Haocong Cheng, Ke-Rou Wang, and Yun Huang. 2021. Examining interactions between community members and university safety organizations through community-sourced risk systems. Proceedings of the ACM on Human-Computer Interaction 5, CSCW1 (2021), 1–23

    2021

  27. [35]

    Isabella Paoletti. 2012. Operators managing callers’ sense of urgency in calls to the medical emergency number. Pragmatics 22, 4 (2012), 671–695

    2012

  28. [36]

    Marieke Peeters, Karel van den Bosch, John-Jules Ch Meyer, and Mark A Neerincx. 2014. The design and effect of automated directions during scenario-based training. Computers & Education 70 (2014), 173–183

    2014

  29. [37]

    Christian Reuter, Volkmar Pipek, and Claudia Mueller. 2009. Computer supported collaborative training in crisis communication management. In Proceedings of the 6th International International Conference on Information Systems for Crisis Response and Management (ISCRAM 2009), Göteburg

    2009

  30. [38]

    Christian Reuter, Volkmar Pipek, and Claudia Muller. 2009. Avoiding crisis in communication: a computer-supported training approach for emergency management. International Journal of Emergency Management 6, 3-4 (2009), 356–368

    2009

  31. [39]

    Eduardo Salas, Heather A Priest, Katherine A Wilson, and C Shawn Burke. 2006. Scenario-Based Training: Improving Military Mission Performance and Adaptability. (2006)

    2006

  32. [40]

    Leonard Salewski, Stephan Alaniz, Isabel Rio-Torto, Eric Schulz, and Zeynep Akata. 2024. In-Context Impersonation Reveals Large Language Models’ Strengths and Biases. Advances in Neural Information Processing Systems 36 (2024)

    2024

  33. [41]

    Alexandre Salle, Shervin Malmasi, Oleg Rokhlenko, and Eugene Agichtein. 2021. Studying the Effectiveness of Conversational Search Refinement Through User Simulation. In Advances in Information Retrieval: 43rd European Conference on IR Research, ECIR 2021, Virtual Event, March ...

    2021 doi

  34. [42]

    Kristen M Scott, Sonja Mei Wang, Milagros Miceli, Pieter Delobelle, Karolina Sztandar-Sztanderska, and Bettina Berendt. 2022. Algorithmic tools in public employment services: Towards a jobseeker-centric perspective. In Proceedings of the 2022 ACM Conference on Fairness, Accoun...

    2022

  35. [43]

    Ivan Sekulić, Silvia Terragni, Victor Guimarães, Nghia Khau, Bruna Guedes, Modestas Filipavicius, André Ferreira Manso, and Roland Mathis. 2024. Reliable LLM-based User Simulator for Task-Oriented Dialogue Systems. arXiv preprint arXiv:2402.13374 (2024)

    2024 arXiv

  36. [44]

    Hannah Shaw and Minna Lyons. 2017. Lie detection accuracy—the role of age and the use of emotions as a reliable cue. Journal of Police and Criminal Psychology 32 (2017), 300–304

    2017

  37. [45]

    Felix Stahlberg and Shankar Kumar. 2021. Synthetic Data Generation for Grammatical Error Correction with Tagged Corruption Models. In Proceedings of the 16th Workshop on Innovative Use of NLP for Building Educational Applications . Association for Computational Linguistics, On...

    2021

  38. [46]

    Mohammed Tahri Sqalli, Dena Al-Thani, Mohamed B Elshazly, and Mohammed Al-Hijji. 2006. Interpretation of a 12-lead electrocardiogram by medical students: quantitative eye-tracking approach. (2006)

    2006

  39. [47]

    Zhaoxuan Tan and Meng Jiang. 2023. User Modeling in the Era of Large Language Models: Current Research and Future Directions. arXiv preprint arXiv:2312.11518 (2023)

    2023 arXiv

  40. [48]

    Silvia Terragni, Modestas Filipavicius, Nghia Khau, Bruna Guedes, André Manso, and Roland Mathis. 2023. In-Context Learning User Simulators for Task-Oriented Dialog Systems. arXiv preprint arXiv:2306.00774 (2023)

    2023 arXiv

  41. [49]

    SM Tonmoy, SM Zaman, Vinija Jain, Anku Rani, Vipula Rawte, Aman Chadha, and Amitava Das. 2024. A comprehensive survey of hallucination mitigation techniques in large language models. arXiv preprint arXiv:2401.01313 (2024)

    2024 arXiv

  42. [50]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288 (2023)

    2023 arXiv

  43. [51]

    Sarah J Tracy and Karen Tracy. 1998. Emotion labor at 911: A case study and theoretical critique. (1998)

    1998

  44. [52]

    Rhema Vaithianathan, Emily Putnam-Hornstein, Nan Jiang, Parma Nand, and Tim Maloney. 2017. Developing predictive models to support child maltreatment hotline screening decisions: Allegheny County methodology and implementation. Center for Social data Analytics (2017)

    2017

  45. [53]

    Nagagopiraju Vullam, Sai Srinivas Vellela, Venkateswara Reddy, M Venkateswara Rao, Khader Basha SK, and D Roja. 2023. Multi-Agent Personalized Recommendation System in E-Commerce based on User. In 2023 2nd International Conference on Applied Artificial Intelligence and Computi...

    2023

  46. [54]

    Hongwei Wang, Fuzheng Zhang, Jialin Wang, Miao Zhao, Wenjie Li, Xing Xie, and Minyi Guo. 2019. Exploring high-order user preference on the knowledge graph for recommender systems. ACM Transactions on Information Systems (TOIS) 37, 3 (2019), 1–26

    2019

  47. [55]

    Gemma Warren, Elizabeth Schertler, and Peter Bull. 2009. Detecting deception from emotional and unemotional cues. Journal of Nonverbal Behavior 33 (2009), 59–69

    2009

  48. [56]

    Hysteria

    Jack Whalen and Don H. Zimmerman. 1998. Observations on the Display and Management of Emotion in Naturally Occurring Activities: The Case of "Hysteria" in Calls to 9-1-1. Social Psychology Quarterly 61, 2 (1998), 141–159. http://www.jstor.org/stable/2787066

    1998

  49. [57]

    Kyohei Yamashita, Yukiko Kato, Koichi Kurabe, Motoki Koike, Koji Jinno, Kazumasa Kito, Kyoichi Tatsuno, and Mohammed Tarhri Sqalli. 2016. Remote operation of a robot for maintaining electric power distribution system using a joystick and a master arm as a human robot interface...

    2016

  50. [58]

    Lantao Yu, Weinan Zhang, Jun Wang, and Yong Yu. 2017. Seqgan: Sequence generative adversarial nets with policy gradient. In Proceedings of the AAAI conference on artificial intelligence , Vol. 31

    2017

  51. [59]

    Yizhe Zhang, Zhe Gan, Kai Fan, Zhi Chen, Ricardo Henao, Dinghan Shen, and Lawrence Carin. 2017. Adversarial feature matching for text generation. In International conference on machine learning . PMLR, 4006–4015

    2017

  52. [60]

    Cheng Zhou and Qian Chang. 2024. Informational or emotional? Exploring the relative effects of chatbots’ self-recovery strategies on consumer satisfaction. Journal of Retailing and Consumer Services 78 (2024), 103779. Received 20 February 2007; revised 12 March 2009; accepted ...

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.