Pith. sign in

REVIEW 4 major objections 6 minor 78 references

Facilitating Video Story Interaction with Multi-Agent Collaborative System

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read A multi-agent system lets viewers chat with video-story characters that appear to grow and evolve across story stages.

desk verdict A genuinely useful design space and prototype for interactive video storytelling, but the headline 'emergent character growth' claim is largely a product of the stage-dependent retrieval design and needs a control condition before it can be taken at face value. read the letter →

arxiv 2505.03807 v1 pith:JEHYHUUC submitted 2025-05-02 cs.HC cs.AIcs.CVcs.MA

classification cs.HCcs.AIcs.CVcs.MA
keywords videostoryinteractionmulti-agentsystemretrieval-augmentedgenerationvisionlanguagemodelcharactergrowthinteractivenarrativescenecustomizationHarryPotter
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

This paper claims that a fixed, pre-shot video story can support genuinely interactive, personalized exploration if the machine first understands the story across vision, audio, and plot, and then lets users talk with characters who are driven by a multi-agent system grounded in the current narrative stage. The intended payoff is that characters show growth potential: the same question asked at an early stage and a late stage gets answers reflecting different knowledge, self-concept, and relationships, so users feel they are walking alongside evolving people rather than selecting pre-written branches. The paper reports a formative interview study with 12 professionals, a system applied to the first two Harry Potter movies, and a user study with 18 participants in which the growth-capable characters and scene customization were rated the most useful functions. This matters because interactive video so far has been limited to choice trees and specially designed narratives; the paper's route would in principle work on existing films.

What carries the argument

The load-bearing mechanism is the stage-conditioned generation loop defined by Equations 3 and 4: a RAGAgent computes stage-specific context $C_{stage_i}$ as a chunked retrieval from vision, audio, and plot text, then each character agent produces $O_{character_i} = LLM(C_{stage_i}+Q_{User}+P_{character_i}+I_{MAS_i})$. The interaction memory stream records each user query and each character's reply so that later rounds are conditioned on prior dialogue. The growth potential is conceptually defined by Equation 2, $d(growth)=f'(stage)\cdot d(stage)$, making character change a function of narrative stage by construction.

What would settle it

A reader could test this by asking the same question at two stages and swapping the retrieved stage contexts: if responses still look stage-coherent, or if users cannot identify which stage a response came from, then the growth effect is driven by retrieval prompts rather than by the characters' accumulated narrative state.

Watch

Extended reading notes

Core claim

The central claim is that retrieval-augmented generation combined with a multi-agent system lets a fixed video story support open-ended, stage-aware dialogue with characters that appear to grow. Each narrative stage has its own interaction space; a RAGAgent retrieves stage-specific chunks of plot, dialogue, and visual description for the user's question, and character agents answer from that context, so the same question asked in stage 1 and stage 5 draws on different knowledge and yields different self-descriptions. The paper interprets these differences as emergent social behavior and growth, using objective self-awareness theory to classify them as knowledge accumulation, awareness of others, self-awareness, and relationship awareness. It reports that 18 users rated the character-growth and scene-customization modules highest and that participants familiar with the story observed stage-consistent changes in Harry's and Hermione's responses.

Load-bearing premise

The load-bearing premise is that the stage-dependent variation in the characters' answers is genuine evidence of psychological growth rather than an effect of swapping which story facts are placed in front of the language model.

Editorial extensions

If this is right

  • Open-ended questions can replace fixed choice points: users ask what they want, and the RAGAgent retrieves the stage-specific context needed for an in-character answer.
  • Characters diverge: the same question produces different answers from the same character at different stages and from different characters at the same stage, which users read as growth and personality.
  • Dialogue can drive visualization: a scene mentioned in chat can be expanded, re-perspectived, or given biography in WebVR, so the user's own assumptions become visible.
  • The approach works on existing films with fixed plots, such as Harry Potter, without rewriting the narrative.
  • Stage-wise interaction spaces let users experience the story as a progression and notice changes between stages, which was the most highly rated function in the user study.

Reading between the lines

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

  • If the growth effect is mostly a retrieval effect, the architecture should reproduce perceived character growth on any serialized story, so a direct test is to run the same design on a non-Harry-Potter corpus and measure whether users report growth without the authors' chosen story.
  • A stronger causal test than the paper reports would scramble stage labels: asking the same question with stage-1 context mislabeled as stage-5 and vice versa, then asking users to judge which stage each response belongs to.
  • The interaction memory stream is session-local; the paper does not give characters persistent memory of one user across days, so a natural extension is to make growth user-specific: a character that remembers prior conversations with the same viewer and changes accordingly.
  • The same stage-conditioned character design could transfer to educational or biographical interaction, where 'stage' is a curriculum or a life period, though the paper only gestures at children's learning in the discussion.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper describes an interactive video-story system that combines a vision-language model (VLM) for video understanding, retrieval-augmented generation (RAG), and a multi-agent system (MAS) to let users chat with characters at different narrative stages and to customize story scenes. The system is applied to the first two Harry Potter films, with narrative data divided into five stages. The authors report a formative interview study with 12 experts, a user study with 18 participants, and case examples intended to show emergent character growth and social behavior. The central claim, stated in the abstract and conclusion, is that the system effectively portrays emergent character social behavior and growth and enhances the interactive experience.

Significance. If the central claim were fully supported, the system would make a useful contribution to interactive storytelling and human-computer interaction by integrating machine video understanding, retrieval-augmented generation, and multi-agent role-play in a single workflow. The formative study, the concrete system implementation, and the qualitative user feedback are valuable design resources, and the paper is commendably explicit about several limitations, including model hallucination, limited VLM visual understanding, and the need for further work on perceived character growth. However, the evidence presented does not yet distinguish system-designed stage-dependent retrieval from genuine, emergent character development, which is the load-bearing part of the claimed contribution.

major comments (4)
  1. [Section 5.2.2, Eqs. (2)-(4)] The claim of emergent character growth is largely guaranteed by construction. Equation 3 defines the context C_stage_i as chunks retrieved for a specific story stage, Equation 4 makes the character output O_character_i depend directly on C_stage_i, and Equation 2 defines d(growth) = f'(stage) * d(stage). Any change in retrieved context therefore produces a change in output that the paper labels as growth. The case study in Section 6.2 (Figure 10), such as the stage-1 versus stage-5 answer to "Who is Dobby?", demonstrates knowledge accumulation in the retrieval corpus, not necessarily psychological development of the character. To support the abstract and conclusion claims, the authors should add a control condition that separates retrieval-driven knowledge updating from character-level developmental change, for example by comparing against a no-RAG condition, a shuffled-stage retrieval condition, or repeated questions within the same stage.
  2. [Section 7.2-7.3, Table 1 and Figure 15] The user evaluation does not include any baseline condition or statistical comparison. With 18 participants, a 7-point Likert questionnaire, and follow-up interviews, the ratings shown in Figures 14-15 can support qualitative usability insights but cannot by themselves establish that the system "effectively portrays emergent character social behavior and growth." The paper also reports in Section 7.3 that model hallucination tends to make character responses converge or include irrelevant information, and Figure 15 shows a large standard deviation for the "difference in persona of different characters is noticeable" item. These observations weaken the strong claim and should be addressed either with additional comparative data or with a more circumscribed statement of what the user study actually demonstrates.
  3. [Section 7.4.1] The paper's own analysis shows that perceived character growth depends on the user's prior familiarity with the story and on the order of stages experienced. Specifically, users who had only read part of the story "often struggle to discern the characters' growth," users who had never seen the film mainly perceive knowledge accumulation, and users who were almost completely familiar perceive self-awareness and other-awareness. This indicates that the reported growth may be substantially driven by the user's external knowledge and the stage-ordering of the interface, rather than by an intrinsic, emergent property of the agent. The authors should either control for familiarity and stage order in the analysis or qualify the growth claim accordingly.
  4. [Section 6.2] The classification of character responses into knowledge accumulation, other-awareness, self-awareness, and relationship awareness is presented as evidence of social behavior, but the coding procedure is not described. No coding scheme, inter-rater reliability measure, or systematic sampling method is given, so it is unclear how the categories were applied and whether they are reproducible. Since this classification is the main evidence for the "social behavior" component of the central claim, the authors should provide a transparent coding protocol or reframe the four categories as illustrative observations rather than measured outcomes.
minor comments (6)
  1. [Section 4.2] There is a grammatical error in the sentence "The dialogue extracted from raw movie By applying the 'moviepy' library," and the audio extraction step should be described in a complete, proofread sentence.
  2. [Section 3.2] Several bullets under "Task" contain incomplete expressions with missing symbols, such as "Characters & Conflict + ," and "Plot + ," which makes the mapping between tasks and narrative elements difficult to follow. These should be fixed before publication.
  3. [Section 5.2.2, Eq. (2)] The function f(stage) appearing in Eq. (2) is not formally defined. The authors should state what f represents, how it is computed, and whether it is a theoretical device or an implemented quantity in the system.
  4. [Section 5.2.2, Eq. (3)] Equation 3 writes C_stage_i as a sum over Chunk_j but does not specify the range of j or the relationship between chunks and stages. Clarifying this notation would help readers understand how retrieval exactly selects stage-specific context.
  5. [Figure 15] The text says the right side shows mean and standard deviation for each question, but the figure does not display numerical values or error bars, and the main text does not report the actual means and standard deviations. Reporting these numbers in a table would make the evaluation results concrete.
  6. [Section 8.1.2] The sentence "all 8 users mentioned" appears inconsistent with the stated participant count of 18. Please check whether this refers to a subset and clarify the wording.

Circularity Check

2 steps flagged · score 7.0 of 10

The central 'emergent character growth' claim reduces by construction to stage-dependent RAG retrieval: Eq. (2) defines growth as a function of stage, Eqs. (3)-(4) make every character response a function of the stage-specific context, and the Section 6.2 case study then exhibits the stage differences the architecture guarantees.

  1. self definitional [Section 5.2.2, Equations (2)-(4); Section 6.2, Figure 10]
    "Definition. We define growth potential as a continuous function that evolves with the video’s narrative stages, represented by Equation 2. ... 𝑑(𝑔𝑟𝑜𝑤𝑡ℎ) = 𝑓′(𝑠𝑡𝑎𝑔𝑒)· 𝑑(𝑠𝑡𝑎𝑔𝑒) (2) ... 𝐶𝑠𝑡𝑎𝑔𝑒 𝑖 = ∑Chunk𝑗 (3) 𝑂𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟 𝑖 = LLM(𝐶𝑠𝑡𝑎𝑔𝑒 𝑖+𝑄𝑈𝑠𝑒𝑟 +𝑃𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟 𝑖+𝐼𝑀𝐴𝑆𝑖) (4)"

    Equation 2 defines character growth as a function of narrative stage, and Equations 3-4 make the retrieved context, and therefore every character response, stage-dependent by construction. The Section 6.2 case study then shows exactly that stage-dependence: asking 'Who is Dobby?' yields 'We haven't encountered him yet' at Stage 1 and a detailed house-elf biography at Stage 5 because the RAGAgent retrieves different stage-specific chunks. That variation is guaranteed by the architecture and demonstrates knowledge accumulation in the retrieval corpus, not a separately observed emergent psychological development. Labeling the guaranteed variation 'growth' and offering it as evidence of emergent behavior is a definitional consequence of the design, not an independent result.

  2. renaming known result [Section 6.2, paragraphs after Figure 10; Section 7.3, Character paragraph]
    "Drawing on Objective Self-Awareness Theory [35, 57] from Social and Personality Psychology, we classified these behaviors into four categories. ... In comparing (2a) and (2b), characters evolve in their understanding of others’ intentions and emotions. ... Almost all users observed that the characters grew as the story progressed through different types of questions, including knowledge growth and self-awareness, which we will discuss in detail in section 7.4.1."

    The four categories in Section 6.2 are labels applied to exactly the variation Equations 3-4 produce: stage-specific retrieval changes the information available to the LLM, and the outputs differ accordingly. The 'evolve' language in the case study adds no evidence beyond that context swap; it renames architectural stage-dependence as psychological growth. The user-study observation cited in Section 7.3 is likewise the perception of the built-in stage-conditioning, measured without a baseline such as no-RAG, shuffled-stage, or repeated same-stage questions, so the reported 'growth' is a perception check of the design rather than validation of emergence. The paper's own Section 8.3.3 concedes that 'the user's perception of character growth still needs attention.'

full rationale

The paper is largely self-contained and does not rely on any load-bearing self-citation chain or imported uniqueness theorem; the formative interviews, RAG pipeline, and scene-customization contributions are independent of the circularity concern. However, the headline claim of 'emergent character social behavior and growth' is partially circular by construction. Equation 2 defines d(growth) as a function of stage, Equations 3-4 make the retrieved context and hence every character response stage-dependent, and Section 6.2's cases illustrate exactly the variation the architecture guarantees. The user study's growth ratings measure users' perception of this engineered stage-dependence, with no baseline that could distinguish retrieval-content accumulation from psychological growth, and the paper itself flags in Section 8.3.3 that perception of character growth still needs attention. The scene customization and general system-usability contributions are not circular, but the central emergence claim is, in its current form, a designed consequence rather than an independently observed phenomenon. Score 7 reflects this partial but central circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 4 invented entities

All component techniques are standard, but the system-level claims of growth and emergence depend on several domain assumptions, including the validity of narrative theory and social psychology categories applied to LLM text. No free parameters are fitted to external benchmarks; the stage divisions, prompt templates, and retrieval settings are hand-chosen. The introduced agents and interaction spaces are software constructs without independent falsifiable evidence beyond the in-paper user study.

free parameters (4)
  • Story stage divisions = 5 stages hand-selected from Harry Potter movies 1-2
    Growth comparisons depend entirely on where stages are cut; a different division would produce different apparent growth curves.
  • Keyframe sampling interval = Not specified
    Controls which visual details the VLM sees and therefore the story context returned by RAG.
  • RAG chunk count and retrieval threshold = Not specified
    Determines how much stage context biases each character response; the paper reports only that the most relevant chunk is retrieved.
  • Agent prompt templates = Bespoke, not released
    Character voice, stage alignment, and first-speaker selection are all driven by hand-written instructions that are not included.
assumptions (5)
  • domain assumption The six-elements narrative theory (plot, theme, setting, character, perspective, conflict) is an adequate basis for structuring user intentions in video story interaction.
    Used to design the task taxonomy in Section 3.2; if this theory mischaracterizes user needs, the design space is invalid.
  • domain assumption Objective Self-Awareness Theory categories can be applied to classify LLM-generated character responses.
    Section 6.2 maps chatbot outputs to knowledge accumulation, other-awareness, self-awareness, and relationship awareness without empirical validation of this mapping.
  • domain assumption GPT-4-vision descriptions of keyframes are sufficiently accurate for story understanding.
    Section 4.2 relies on GPT-4-vision; Section 8.3.3 admits it is limited to basic visual descriptions and cannot understand complete specific characters and scenes.
  • standard math Embedding similarity retrieval returns context relevant to user intent.
    The RAGAgent uses embedding distances (Section 5.2.2); this is a standard assumption of vector retrieval systems.
  • domain assumption LLM role-play output reliably reflects the target character's personality at a given story stage.
    The whole system is built on this; the paper reports hallucination and response convergence issues in Section 7.3.
invented entities (4)
  • RAGAgent
    purpose: Retrieves stage-specific story context and coordinates which character responds first.
    It is a core software component of the system; its effectiveness is shown only through the in-paper user study, with no separate benchmark.
  • Character Agent with Growth Potential
    purpose: Produces stage-aware character responses in multi-user chat.
    The growth behavior is designed via per-stage context retrieval, so it cannot falsify the growth claim independently.
  • Interaction Space
    purpose: Links video story stages to chat and scene-customization environments.
    It is a UI concept inspired by coroutines; no independent validation outside the user study.
  • Trans-Temporal Sharing
    purpose: Generates character reflections and shared objects based on user focus and mood.
    Outputs are generated by the same LLM pipeline; no external benchmark measures their accuracy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Facilitating Video Story Interaction with Multi-Agent Collaborative System." pith.science (2026). https://pith.science/paper/JEHYHUUC

@misc{pith2026250503807,
  author       = {Pith},
  title        = {Pith review of: Facilitating Video Story Interaction with Multi-Agent Collaborative System},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JEHYHUUC}},
  note         = {Machine review of arXiv:2505.03807}
}
read the original abstract

Video story interaction enables viewers to engage with and explore narrative content for personalized experiences. However, existing methods are limited to user selection, specially designed narratives, and lack customization. To address this, we propose an interactive system based on user intent. Our system uses a Vision Language Model (VLM) to enable machines to understand video stories, combining Retrieval-Augmented Generation (RAG) and a Multi-Agent System (MAS) to create evolving characters and scene experiences. It includes three stages: 1) Video story processing, utilizing VLM and prior knowledge to simulate human understanding of stories across three modalities. 2) Multi-space chat, creating growth-oriented characters through MAS interactions based on user queries and story stages. 3) Scene customization, expanding and visualizing various story scenes mentioned in dialogue. Applied to the Harry Potter series, our study shows the system effectively portrays emergent character social behavior and growth, enhancing the interactive experience in the video story world.

Figures

Figures reproduced from arXiv: 2505.03807 by the authors.

Figure 1
Figure 1. Design framework for video story interaction using MAS. Based on a semi-structured interview in the formative study and the [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. System overview. The system involves a machine understanding a video story and incorporating that understanding into [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. VLM-based video story comprehension pipeline. The use of a Vision-Language Model (VLM) allows machines to comprehend [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: The input and output utilize MAS as the main component. Once the machine comprehends the video story, MAS handles and [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Multi space chat design concept. The integration of fluid narrative and high-freedom interactive spaces using the startCoroutine( [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: The two-round response generation process for characters in response to user queries at different stages of a story. User queries are matched to the relevant stage information in the video story dataset through an embedding process. The retrieved text serves as context…
Figure 7
Figure 7. Figure 7: The single-round query operation within each interactive space, from user query to character response, integrates RAG technology with MAS to enable character agents to provide rich, accurate answers from their perspectives. Technical Comparison. There are three main te…
Figure 8
Figure 8. Figure 8: Conceptual model illustrates the technical comparison. The effectiveness of text input and in-context learning depends on [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: User interface. The user interaction functions are displayed across distinct zones: (1) Video Story Stage Selection, (2) Trans [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: Cases of the agent with growth potential. (3a) and (4a) represent Stage 1, (1a) and (2a) represent Stage 2, and the remaining [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Different characters provide different perspectives. (5a) presents simultaneous responses from Harry and Voldemort in Stage [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: The cross-temporal sharing that the characters discuss after interacting with users. The two main identifications are User [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: The customized scenes and their descriptions after two rounds of questions with different characters, using three methods. [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]
Figure 14
Figure 14. Figure 14: The usefulness rating results of our unique new functions. Scene Customization and Characters with Growth Potential were [PITH_FULL_IMAGE:figures/full_fig_p021_14.png]
Figure 15
Figure 15. Figure 15: The evaluation results of our system, contain three aspects: the overall evaluation of the interactive system, the evaluation of [PITH_FULL_IMAGE:figures/full_fig_p022_15.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

78 extracted references · 45 canonical work pages

  1. [1]

    ‘What If

    2024. ‘What If. . . ? – An Immersive Story’ Release Date & Trailer Revealed. https://www.marvel.com/articles/games/what-if-an-immersive-story- release-date-trailer

  2. [2]

    Michael Ahn, Debidatta Dwibedi, Chelsea Finn, Montse Gonzalez Arenas, Keerthana Gopalakrishnan, Karol Hausman, Brian Ichter, Alex Irpan, Nikhil Joshi, Ryan Julian, et al. 2024. Autort: Embodied foundation models for large scale orchestration of robotic agents. arXiv preprint arXiv:2401.12963 (2024)

  3. [3]

    Jens B Asendorpf and Pierre-Marie Baudonnière. 1993. Self-awareness and other-awareness: Mirror self-recognition and synchronic imitation among unfamiliar peers. Developmental Psychology 29, 1 (1993), 88

  4. [4]

    Craig Caldwell. 2020. What we talk about, when we talk about story. In ACM SIGGRAPH 2020 Courses. 1–15

  5. [5]

    Monica Callahan. 2019. Presenting foreshadowing

  6. [6]

    Edwin Catmull. 2008. How Pixar fosters collective creativity . Harvard Business School Publishing Boston, MA

  7. [7]

    Chi-Min Chan, Weize Chen, Yusheng Su, Jianxuan Yu, Wei Xue, Shanghang Zhang, Jie Fu, and Zhiyuan Liu. [n. d.]. ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate. In The Twelfth International Conference on Learning Representations

  8. [8]

    Dake Chen, Hanbin Wang, Yunhao Huo, Yuzhao Li, and Haoyang Zhang. 2023. Gamegpt: Multi-agent collaborative framework for game development. arXiv preprint arXiv:2310.08067 (2023)

Show all 78 references
  1. [9]

    Jenova Chen. 2007. Flow in games (and everything else). Commun. ACM 50, 4 (2007), 31–34

  2. [10]

    Nuo Chen, Yan Wang, Haiyun Jiang, Deng Cai, Yuhan Li, Ziyang Chen, Longyue Wang, and Jia Li. 2022. Large Language Models Meet Harry Potter: A Bilingual Dataset for Aligning Dialogue Agents with Characters. arXiv preprint arXiv:2211.06869 (2022)

  3. [11]

    Weize Chen, Yusheng Su, Jingwei Zuo, Cheng Yang, Chenfei Yuan, Chen Qian, Chi-Min Chan, Yujia Qin, Yaxi Lu, Ruobing Xie, et al. 2023. Agentverse: Facilitating multi-agent collaboration and exploring emergent behaviors in agents. arXiv preprint arXiv:2308.10848 (2023)

  4. [12]

    Yuheng Cheng, Ceyao Zhang, Zhengwen Zhang, Xiangrui Meng, Sirui Hong, Wenhao Li, Zihao Wang, Zekai Wang, Feng Yin, Junhua Zhao, et al

  5. [13]

    Seongho Choi, Kyoung-Woon On, Yu-Jung Heo, Ahjeong Seo, Youwon Jang, Minsu Lee, and Byoung-Tak Zhang. 2021. Dramaqa: Character-centered video story understanding with hierarchical qa. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 35. 1166–1174. Manusc...

  6. [14]

    Bhavya Chopra, Khushali Verma, Sonali Singhal, and Utsav Singla. 2021. Reality Tales: Facilitating User-Character Interaction with Immersive Storytelling. In Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing Systems . 1–7

  7. [15]

    Yun-Shiuan Chuang, Nikunj Harlalka, Siddharth Suresh, Agam Goyal, Robert Hawkins, Sijia Yang, Dhavan Shah, Junjie Hu, and Timothy T Rogers

  8. [16]

    Mihaly Czikszentmihalyi. 1990. Flow: The psychology of optimal experience . New York: Harper & Row

  9. [17]

    In Proceedings of the Annual Meeting of the Cognitive Science Society , Vol

    The Wisdom of Partisan Crowds: Comparing Collective Intelligence in Humans and LLM-based Agents. In Proceedings of the Annual Meeting of the Cognitive Science Society , Vol. 46

  10. [18]

    Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Zhiyong Wu, Baobao Chang, Xu Sun, Jingjing Xu, and Zhifang Sui. 2022. A survey on in-context learning. arXiv preprint arXiv:2301.00234 (2022)

  11. [19]

    Sohail Dahdal. 2020. The illusive ludonarrativity and the problem with emergent interactive storytelling models in interactive movies. Journal of Digital Media & Interaction 3, 6 (2020), 17–33

  12. [20]

    Zane Durante, Qiuyuan Huang, Naoki Wake, Ran Gong, Jae Sung Park, Bidipta Sarkar, Rohan Taori, Yusuke Noda, Demetri Terzopoulos, Yejin Choi, et al. 2024. Agent ai: Surveying the horizons of multimodal interaction. arXiv preprint arXiv:2401.03568 (2024)

  13. [21]

    Yihong Dong, Xue Jiang, Zhi Jin, and Ge Li. 2023. Self-collaboration code generation via chatgpt. arXiv preprint arXiv:2304.07590 (2023)

  14. [22]

    Pablo Gervás. 2009. Computational approaches to storytelling and creativity. AI Magazine 30, 3 (2009), 49–49

  15. [23]

    Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, and Haofen Wang. 2023. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997 (2023)

  16. [24]

    Sun Hang and Mustaffa Halabi Azahari. 2023. Interactive movie in digital era: A systematic literature review. (2023)

  17. [25]

    Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V Chawla, Olaf Wiest, and Xiangliang Zhang. 2024. Large language model based multi-agents: A survey of progress and challenges. arXiv preprint arXiv:2402.01680 (2024)

  18. [26]

    Wenyue Hua, Lizhou Fan, Lingyao Li, Kai Mei, Jianchao Ji, Yingqiang Ge, Libby Hemphill, and Yongfeng Zhang. 2023. War and peace (waragent): Large language model-based multi-agent simulation of world wars. arXiv preprint arXiv:2311.17227 (2023)

  19. [27]

    Sirui Hong, Xiawu Zheng, Jonathan Chen, Yuheng Cheng, Jinlin Wang, Ceyao Zhang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, Liyang Zhou, et al. 2023. Metagpt: Meta programming for multi-agent collaborative framework. arXiv preprint arXiv:2308.00352 (2023)

  20. [28]

    Qingqiu Huang, Yu Xiong, Anyi Rao, Jiaze Wang, and Dahua Lin. 2020. Movienet: A holistic dataset for movie understanding. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IV 16 . Springer, 709–727

  21. [29]

    Qiuyuan Huang, Naoki Wake, Bidipta Sarkar, Zane Durante, Ran Gong, Rohan Taori, Yusuke Noda, Demetri Terzopoulos, Noboru Kuno, Ade Famoti, et al. 2024. Position Paper: Agent AI Towards a Holistic Intelligence. arXiv preprint arXiv:2403.00833 (2024)

  22. [30]

    Afzal Hussain, Haad Shakeel, Faizan Hussain, Nasir Uddin, and Turab Latif Ghouri. 2020. Unity game development engine: A technical survey. Univ. Sindh J. Inf. Commun. Technol 4, 2 (2020), 73–81

  23. [31]

    Yu Huang. 2024. Levels of AI Agents: from Rules to Large Language Models. arXiv preprint arXiv:2405.06643 (2024)

  24. [32]

    Rensis Likert. 1932. A technique for the measurement of attitudes. Archives of psychology (1932)

  25. [33]

    Tian Liang, Zhiwei He, Jen-tes Huang, Wenxuan Wang, Wenxiang Jiao, Rui Wang, Yujiu Yang, Zhaopeng Tu, Shuming Shi, and Xing Wang. 2023. Leveraging Word Guessing Games to Assess the Intelligence of Large Language Models. arXiv preprint arXiv:2310.20499 (2023)

  26. [34]

    Chang Liu, Armin Shmilovici, and Mark Last. 2020. Towards story-based classification of movie scenes. PloS one 15, 2 (2020), e0228579

  27. [35]

    Shuhang Lin, Wenyue Hua, Lingyao Li, Che-Jui Chang, Lizhou Fan, Jianchao Ji, Hang Hua, Mingyu Jin, Jiebo Luo, and Yongfeng Zhang. 2024. BattleAgent: Multi-modal Dynamic Emulation on Historical Battles to Complement Historical Analysis. arXiv preprint arXiv:2404.15532 (2024)

  28. [36]

    Jiaying Lu, Bo Pan, Jieyi Chen, Yingchaojie Feng, Jingyuan Hu, Yuchen Peng, and Wei Chen. 2024. AgentLens: Visual Analysis for Agent Behaviors in LLM-based Autonomous Systems. arXiv preprint arXiv:2402.08995 (2024)

  29. [37]

    Manuel London, Valerie I Sessa, and Loren A Shelley. 2023. Developing self-awareness: Learning processes for self-and interpersonal growth. Annual Review of Organizational Psychology and Organizational Behavior 10 (2023), 261–288

  30. [38]

    Weiyu Ma, Qirui Mi, Xue Yan, Yuqiao Wu, Runji Lin, Haifeng Zhang, and Jun Wang. 2023. Large language models play starcraft ii: Benchmarks and a chain of summarization approach. arXiv preprint arXiv:2312.11865 (2023)

  31. [39]

    Lynshao Celina Ma. 2023. Enriched Story Experiences with a New Video Interaction Model. (2023)

  32. [40]

    Tula Masterman, Sandi Besen, Mason Sawtell, and Alex Chao. 2024. The Landscape of Emerging AI Agent Architectures for Reasoning, Planning, and Tool Calling: A Survey. arXiv preprint arXiv:2404.11584 (2024)

  33. [41]

    David March, Julia Múgica, Ezequiel E Ferrero, and M Carmen Miguel. 2023. Honeybee-like collective decision making in a kilobot swarm. arXiv preprint arXiv:2310.15592 (2023)

  34. [42]

    Robert McKee. 1997. Substance, structure, style, and the principles of screenwriting. Alba Editorial (1997)

  35. [43]

    The road not taken

    Grzegorz Maziarczyk. 2023. “The road not taken”: an interactive film between narrative and database. New Review of Hypermedia and Multimedia 29, 1 (2023), 56–71

  36. [44]

    Stanley Milgram and Christian Gudehus. 1974. Obedience to authority

  37. [45]

    Stanley Milgram. 1963. Behavioral study of obedience. The Journal of abnormal and social psychology 67, 4 (1963), 371

  38. [46]

    Janet H Murray. 2018. Research into interactive digital narrative: a kaleidoscopic view. In Interactive Storytelling: 11th International Conference on Interactive Digital Storytelling, ICIDS 2018, Dublin, Ireland, December 5–8, 2018, Proceedings 11 . Springer, 3–17. Manuscript...

  39. [47]

    Shervin Minaee, Tomas Mikolov, Narjes Nikzad, Meysam Chenaghlu, Richard Socher, Xavier Amatriain, and Jianfeng Gao. 2024. Large language models: A survey. arXiv preprint arXiv:2402.06196 (2024)

  40. [48]

    Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. 2023. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology . 1–22

  41. [49]

    Chang-Jun Nan, Kyung-Min Kim, and Byoung-Tak Zhang. 2015. Social network analysis of TV drama characters via deep concept hierarchies. In Proceedings of the 2015 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2015 . 831–836

  42. [50]

    Chen Qian, Wei Liu, Hongzhang Liu, Nuo Chen, Yufan Dang, Jiahao Li, Cheng Yang, Weize Chen, Yusheng Su, Xin Cong, et al. 2024. Chatdev: Communicative agents for software development. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Vo...

  43. [51]

    Celia Pearce. 1994. The ins & outs of non-linear storytelling. ACM SIGGRAPH Computer Graphics 28, 2 (1994), 100–101

  44. [52]

    Mark O Riedl and Robert Michael Young. 2010. Narrative planning: Balancing plot and character. Journal of Artificial Intelligence Research 39 (2010), 217–268

  45. [53]

    Priyanka Ranade and Anupam Joshi. 2023. FABULA: Intelligence Report Generation Using Retrieval-Augmented Narrative Construction. arXiv preprint arXiv:2310.13848 (2023)

  46. [54]

    Marie-Laure Ryan. 2015. Narrative as virtual reality 2: Revisiting immersion and interactivity in literature and electronic media . JHU press

  47. [55]

    Christian Roth and Hartmut Koenitz. 2019. Bandersnatch, yea or nay? Reception and user experience of an interactive digital narrative video. In Proceedings of the 2019 ACM International Conference on Interactive Experiences for TV and Online Video . 247–254

  48. [56]

    Yunfan Shao, Linyang Li, Junqi Dai, and Xipeng Qiu. 2023. Character-LLM: A Trainable Agent for Role-Playing. arXiv:2310.10158 [cs.CL]

  49. [57]

    Claude Elwood Shannon. 1948. A mathematical theory of communication. The Bell system technical journal 27, 3 (1948), 379–423

  50. [58]

    Dushyant Somvanshi. 2023. Audio-Visual Storytelling Through Immersive Media and its Impact: A Critical Literature Review. European Chemical Bulletin (2023)

  51. [59]

    Paul J Silvia and T Shelley Duval. 2001. Objective self-awareness theory: Recent progress and enduring problems. Personality and social psychology review 5, 3 (2001), 230–241

  52. [60]

    Yuzuru Tanahashi and Kwan-Liu Ma. 2012. Design considerations for optimizing storyline visualizations. IEEE Transactions on Visualization and Computer Graphics 18, 12 (2012), 2679–2688

  53. [61]

    Mengdi Sun, Ligan Cai, Weiwei Cui, Yanqiu Wu, Yang Shi, and Nan Cao. 2022. Erato: Cooperative data story editing via fact interpolation. IEEE Transactions on Visualization and Computer Graphics 29, 1 (2022), 983–993

  54. [62]

    Kirti Verma, Sateesh Kourav, Mukul Jangid, Uma Sahu, and Neeraj Shivhare. 2023. Research on Finite State Machine and Its Real Life Time Applications. (12 2023)

  55. [63]

    Marian F Ursu, Maureen Thomas, Ian Kegel, Doug Williams, Mika Tuomola, Inger Lindstedt, Terence Wright, Andra Leurdijk, Vilmos Zsombori, Julia Sussner, et al. 2008. Interactive TV narratives: Opportunities, progress, and challenges. ACM Transactions on Multimedia Computing, Co...

  56. [64]

    Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu, Linxi Fan, and Anima Anandkumar. [n. d.]. Voyager: An Open-Ended Embodied Agent with Large Language Models. Transactions on Machine Learning Research ([n. d.])

  57. [65]

    Paul Vicol, Makarand Tapaswi, Lluis Castrejon, and Sanja Fidler. 2018. Moviegraphs: Towards understanding human-centric situations from videos. In Proceedings of the IEEE conference on computer vision and pattern recognition . 8581–8590

  58. [66]

    Zhan Wang, Lin-Ping Yuan, Liangwei Wang, Bingchuan Jiang, and Wei Zeng. 2024. Virtuwander: Enhancing multi-modal interaction for virtual tour guidance through large language models. In Proceedings of the CHI conference on human factors in computing systems . 1–20

  59. [67]

    Zhenhailong Wang, Shaoguang Mao, Wenshan Wu, Tao Ge, Furu Wei, and Heng Ji. 2023. Unleashing the emergent cognitive synergy in large language models: A task-solving agent through multi-persona self-collaboration. arXiv preprint arXiv:2307.05300 (2023)

  60. [68]

    Patricia H Werhane. 2019. The rashomon effect. Systems thinking and moral imagination: Rethinking Business Ethics with Patricia Werhane (2019), 335–343

  61. [69]

    Chung-Yi Weng, Wei-Ta Chu, and Ja-Ling Wu. 2009. Rolenet: Movie analysis from the perspective of social networks. IEEE Transactions on Multimedia 11, 2 (2009), 256–271

  62. [70]

    Minghao Wu, Yulin Yuan, Gholamreza Haffari, and Longyue Wang. 2024. (Perhaps) Beyond Human Translation: Harnessing Multi-Agent Collaboration for Translating Ultra-Long Literary Texts. arXiv preprint arXiv:2405.11804 (2024)

  63. [71]

    Guande Wu, Shunan Guo, Jane Hoffswell, Gromit Yeuk-Yin Chan, Ryan A Rossi, and Eunyee Koh. 2023. Socrates: Data Story Generation via Adaptive Machine-Guided Elicitation of User Feedback. IEEE Transactions on Visualization and Computer Graphics (2023)

  64. [72]

    Zhiheng Xi, Yiwen Ding, Wenxiang Chen, Boyang Hong, Honglin Guo, Junzhe Wang, Dingwen Yang, Chenyang Liao, Xin Guo, Wei He, Songyang Gao, Lu Chen, Rui Zheng, Yicheng Zou, Tao Gui, Qi Zhang, Xipeng Qiu, Xuanjing Huang, Zuxuan Wu, and Yu-Gang Jiang. 2024. AgentGym: Evolving Larg...

  65. [73]

    Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Shaokun Zhang, Erkang Zhu, Beibin Li, Li Jiang, Xiaoyun Zhang, and Chi Wang. 2023. Autogen: Enabling next-gen llm applications via multi-agent conversation framework. arXiv preprint arXiv:2308.08155 (2023)

  66. [74]

    Xingchen Zeng, Ziyao Gao, Yilin Ye, and Wei Zeng. 2024. IntentTuner: An Interactive Framework for Integrating Human Intentions in Fine-tuning Text-to-Image Generative Models. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–18

  67. [75]

    Qi Xiaoliang. 2024. Time, Information and Artificial Intelligence. Physics 53, 6 (2024), 357–367

  68. [76]

    Junhao Zheng, Shengjie Qiu, Chengming Shi, and Qianli Ma. 2024. Towards Lifelong Learning of Large Language Models: A Survey. arXiv preprint arXiv:2406.06391 (2024). Manuscript submitted to ACM

  69. [77]

    Qinlin Zhao, Jindong Wang, Yixuan Zhang, Yiqiao Jin, Kaijie Zhu, Hao Chen, and Xing Xie. 2023. Competeai: Understanding the competition behaviors in large language model-based agents. arXiv preprint arXiv:2310.17512 (2023). Manuscript submitted to ACM Facilitating Video Story ...

  70. [2024]

    arXiv preprint arXiv:2401.03428 (2024)

    Exploring Large Language Model based Intelligent Agents: Definitions, Methods, and Prospects. arXiv preprint arXiv:2401.03428 (2024)

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.