Jihye Kim (Ji Hye Kim)

Jihye Kim (Ji Hye Kim)

Hi, I am Jihye Kim. I received my Ph.D. in Information Systems from the Department of Management Engineering at KAIST, where I was advised by Prof. Wonseok Oh. I am currently pursuing an additional M.S. in Natural Language Processing in the Department of Computer Science and Engineering at UC Santa Cruz to deepen my technical expertise in large language models and AI agents.

Before my Ph.D., I worked as a Senior Consultant at Deloitte Consulting, where I focused on digital transformation, analytics, and organizational strategy.

Research Themes

Reshaping the Decision Environment

Some technologies shape decisions without making them: they change when people can act, which options become salient, and how choices are made. I study how changes in technological and institutional environments reshape the behavior of workers, consumers, and firms. My work has examined how on-demand wage access changes the way workers manage their finances, how consumers alter their product choices in automated ordering environments, and how firms and consumers respond to shifts in platform rules and governance. To study these questions, I primarily apply econometric methods for causal inference to large-scale behavioral data, examining whose behavior changes, through which mechanisms, and how the resulting benefits and costs are distributed across market participants.

Economics of Information Systems · Digital Platforms · Causal Inference

AI in the Decision Loop

As language models increasingly advise, negotiate, and act on users’ behalf, they become part of the decision-making process rather than passive information sources. I study when LLMs and AI agents fail in these roles and how those failures change as they interact with users, other agents, and system constraints. My work examines whether models can distinguish helpful from misleading signals, how safety interventions change agent behavior and sometimes introduce new problems, and when people become over- or under-reliant on AI. I am especially interested in failures that emerge through interaction or adaptation and are not captured well by standard benchmarks. More broadly, I develop evaluation and mitigation methods to detect and reduce these failures, with the goal of making AI systems more reliable in human and organizational decision-making.

AI Evaluation · Agentic Systems · Human–AI Decision-Making

Papers & Ongoing Work

Published work, working papers, and ongoing projects. * First author; † equal contribution.

Journal Articles

  • Jihye Kim*, Seokchae Yoon, Sunghun Chung, Wonseok Oh . Working Daily, Paid Monthly? Effects of On-Demand Wage Access on the Financial Engagement of Low-Wage Workers . Information Systems Research (UTD24, FT50) · Volume 37, Issue 3 · 2026 . Journal Media

Working Papers & Preprints

  • Jihye Kim*, Seokchae Yoon, Anindya Ghose, Wonseok Oh . Beyond Efficiency: The Impact of Self-Order Kiosk Adoption on Demand Variety . Working paper .
  • Jihye Kim*, Jeffrey Flanigan . Right or Wrong, Models Comply: Directional Blindness in LLM Moral Judgment . arXiv preprint · 2026 . arXiv
  • Shivam Shukla, Jihye Kim†, Shubham Gaur†, Mahnaz Roshanaei, Magy Seif El-Nasr . RELATE: An Evaluation Framework for Measuring Relational Orientation of Large Language Models . arXiv preprint · 2026 . arXiv
  • Jihye Kim*, Umit Azirakhmet, Xuefang Yang . Optimizing Verbal Calibration: Structural Prompting Reducing Confidence Error in LLMs . Working paper .

Refereed Conference & Workshop Papers (Proceedings)

  • Jihye Kim* (Single Author) . Coercion Suppression Increases Preference Hallucinations via a Deceptive Bypass in K-Level Negotiation Agents . 6th Workshop on Trustworthy NLP (TrustNLP) at ACL 2026 . GitHub Paper
  • Pratibha Revankar, Jihye Kim, Umit Azirakhmet . SlugRAG at SemEval-2026 Task 8: Domain-Specific Fine-Tuning and Model Scaling for Multi-Turn RAG Retrieval . SemEval at ACL 2026 . Paper
  • Yoonha Park, Jihye Kim, Kyumin Lee, Wonseok Oh . The Information Billboard: Effects of Popular Search Terms on Search Behaviors and Digital Divide . 57th Hawaii International Conference on System Sciences (HICSS 2024) · Jan 2024 . Paper

Refereed Conference & Workshop Papers (Non-archival)

  • Pratibha Revankar, Kargi Chauhan, Jihye Kim, Sadiba Nusrat Nur, Vincent Siu, Chenguang Wang . MIRAGE: A Polarity-Flipping Encoding Subspace in LLM Agents . 2nd Workshop on Agents in the Wild (AIWILD) at ICML 2026 Paper
  • Jihye Kim*, Seokchae Yoon . Could Self-Order Kiosks Drive Unequal Sales Variety? . 35th Workshop on Information Systems and Economics (WISE 2024) · Bangkok · Dec 2024 Slides
  • Jihye Kim*, Seokchae Yoon, Sunghun Chung . Working Daily, Paid Monthly? Effects of On-Demand Earned Wage Access on the Financial Well-Being of Low-Wage Workers . Conference on Information Systems and Technology (CIST 2023) · Phoenix, AZ · October 2023 Slides
  • Jihye Kim*, Hyeokkoo Eric Kwon, Kun Shin Im, Wonseok Oh . In-App Payment Regulation and Platform Governance . KrAIS Summer Workshop 2023 · Seoul · Jul 2023
  • Jihye Kim*, Seokchae Yoon, Sunghun Chung, Wonseok Oh . On-Demand Earned Wage Access and Financial Well-Being of Low-Wage Workers . Marketing Science: Diversity, Equity & Inclusion Conference (MSI DEI 2023) · Dallas, TX · Mar 2023
  • Jihye Kim*, Hyeokkoo Eric Kwon, Kun Shin Im, Wonseok Oh . The Economics of In-App Payment Options . Conference on Information Systems and Technology (CIST 2022) · Indianapolis, IN · October 2022 . Slides

Research Under Review

To preserve confidentiality during peer review, the entries below use general descriptive labels rather than actual manuscript titles.

Platform Governance and Competition in Digital Markets Under 2nd-round review
Journal of Management Information Systems
Trending Search Features and Online Public Participation Under 2nd-round review
Decision Support Systems

Research in Progress

Benchmarking Data Security Failures in Enterprise Multi-Agent LLM Systems In progress
Ongoing research at UC Santa Cruz in collaboration with Uniphore

Teaching

Courses taught and service

Service: Reviewer for the TAE Workshop at NeurIPS (2026), SemEval at ACL (2026), ICIS (2022–2024), CIST (2023, 2025), and PACIS (2023).

Selected Open-Source Projects

Tools and implementations for working with evidence and extracting information from text.

Sourcerer

Searches a personal PDF library for passages that support a claim, combining semantic retrieval, reranking, and evidence assessment with section-level source references.

Semantic search · Reranking · Evidence attribution

GitHub ↗

Mathematical Entity Extraction

Extracts definitions, theorems, proofs, and examples from mathematics textbooks using a fine-tuned language model, with evaluation against few-shot baselines.

LoRA fine-tuning · Structured extraction · Evaluation

GitHub ↗

Press Coverage

Selected media on my research