arXiv subjects

Jeonghyun Park

Publications and source records attributed to Jeonghyun Park.

11 recordsLinked to original sources

arXiv2026

Identifying and Mitigating Bottlenecks in Role-Playing Agents: A Systematic Study of Disentangling Character Profile Axes

While Large Language Model (LLM) role-playing agents have advanced rapidly, it remains unclear which profile elements genuinely drive role-playing quality. To bridge this gap, we introduce a systematic diagnostic framework that disentangles the impact of character profiles along three axes: Familiarity (Known vs. Unknown), Structure (Structured vs. Unstructured), and Disposition (Moral vs. Immoral). Utilizing a unified hierarchical schema (5 dimensions, 28 fields), we construct a controlled dataset of 211 personas and evaluate five LLMs on both single- and multi-turn interactions. Our results reveal a striking asymmetry: \textbf{Familiarity} and \textbf{Structure} show negligible impact, while \textbf{Disposition} produces large, consistent performance degradation for immoral characters across all conditions. Further analyses suggest that the Moral--Immoral gap is amplified by post-SFT alignment, and that this degradation varies substantially across profile attributes. To mitigate this bottleneck, we propose Field-Aware Contrastive Decoding (FACD), a training-free strategy that amplifies suppressed disposition-sensitive signals, significantly closing the performance gap without sacrificing moral-character performance.

cs.CL↗

arXiv2026

Code-Switching Reveals Language Anchoring in Multilingual LLMs

Multilingual Large Language Models (MLLMs) are increasingly expected to handle Code-Switched (CS) inputs, yet mixing languages frequently degrades performance relative to source- or target-language monolingual counterparts. To understand this degradation, we use grammar-forced CS as a controlled diagnostic setting for locating CS representations relative to their source and target counterparts. We introduce Anchor Bias, a geometric measure that quantifies language anchoring, whether a CS hidden state aligns closer to its source or target language counterpart. Across diverse MLLMs, Anchor Bias reveals a consistent grammar-frame effect: source-framed CS stays source-anchored, whereas target-framed CS shifts target-ward and shows larger Question Answering (QA) degradation. Motivated by this representational pattern, we propose CANVAS (Contextual Anchor-based Neural Vector Alignment Steering), an inference-time intervention that extracts a source-side canvas from the input and softly steers target-language hidden states toward the source anchor during prefill. CANVAS consistently recovers QA F1 across MLLMs and CS conditions, showing that internal anchoring signals provide an actionable target for mitigating CS inference failures.

cs.CL↗

arXiv2026

Beyond One Path: Evaluating and Enhancing Divergent Thinking in Interactive LLM Agents

Divergent thinking is a core dimension of creativity, yet existing evaluations of Large Language Models (LLMs) treat them as single-turn text generations, failing to capture how an agent reasons through iterative interaction. To address this, we introduce MUTATE, an interactive benchmark designed to evaluate agentic divergent thinking at two levels: path-level, where an agent discovers multiple alternative paths to the same goal, and action-level, where individual actions require non-typical, mechanism-shifting object uses. Unlike success-only evaluations, MUTATE scores both completed paths and off-path attempts, capturing divergent reasoning that conventional success rates discard. Our experiments with frontier LLMs reveal a structural blind spot in existing frameworks: when exposed to immediate convergence pressure, they tend to fall into immediate action fixation, failing to improve action-level divergence. To overcome this, we propose ReDNA, which separates unconstrained divergent candidate generation from convergent constraint selection. ReDNA significantly outperforms prior methods across both divergence levels and generalizes effectively to an external creativity environment. We also confirm its success stems from a qualitative enhancement of resilient divergent reasoning rather than simple environmental exploration.

cs.CL↗

arXiv2026

MARCH: Evaluating the Intersection of Ambiguity Interpretation and Multi-hop Inference

Real-world multi-hop QA is naturally linked with ambiguity, where a single query can trigger multiple reasoning paths that require independent resolution. Since ambiguity can occur at any stage, models must navigate layered uncertainty throughout the entire reasoning chain. Despite its prevalence in real-world user queries, previous benchmarks have primarily focused on single-hop ambiguity, leaving the complex interaction between multi-step inference and layered ambiguity underexplored. In this paper, we introduce MARCH, a benchmark for their intersection, with 2,209 multi-hop ambiguous questions curated via multi-LLM verification and validated by human annotation with strong agreement. Our experiments reveal that even state-of-the-art models struggle with MARCH, confirming that combining ambiguity resolution with multi-step reasoning is a significant challenge. To address this, we propose CLARION, a two-stage agentic framework that explicitly decouples ambiguity planning from evidence-driven reasoning, significantly outperforms existing approaches, and paves the way for robust reasoning systems.

cs.CL↗

arXiv2026

Enhancing Multilingual RAG Systems with Debiased Language Preference-Guided Query Fusion

Multilingual Retrieval-Augmented Generation (mRAG) systems often exhibit a perceived preference for high-resource languages, particularly English, resulting in the widespread adoption of English pivoting. While prior studies attribute this advantage to the superior English-centric capabilities of Large Language Models (LLMs), we find that such measurements are significantly distorted by structural priors inherent in evaluation benchmarks. Specifically, we identify exposure bias and a gold availability prior-both driven by the disproportionate concentration of resources in English-as well as cultural priors rooted in topic locality, as factors that hinder accurate assessment of genuine language preference. To address these biases, we propose DeLP (Debiased Language Preference), a calibrated metric designed to explicitly factor out these structural confounds. Our analysis using DeLP reveals that the previously reported English preference is largely a byproduct of evidence distribution rather than an inherent model bias. Instead, we find that retrievers fundamentally favor monolingual alignment between the query and the document language. Building on this insight, we introduce DELTA (DEbiased Language preference-guided Text Augmentation), a lightweight and efficient mRAG framework that strategically leverages monolingual alignment to optimize cross-lingual retrieval and generation. Experimental results demonstrate that DELTA consistently outperforms English pivoting and mRAG baselines across diverse languages.

cs.CL↗

arXiv2025

Chronological Passage Assembling in RAG framework for Temporal Question Answering

Long-context question answering over narrative tasks is challenging because correct answers often hinge on reconstructing a coherent timeline of events while preserving contextual f low in a limited context window. Retrievalaugmented generation (RAG) methods aim to address this challenge by selectively retrieving only necessary document segments. However, narrative texts possess unique characteristics that limit the effectiveness of these existing approaches. Specifically, understanding narrative texts requires more than isolated segments, as the broader context and sequential relationships between segments are crucial for comprehension. To address these limitations, we propose ChronoRAG, a novel RAG framework specialized for narrative texts. This approach focuses on two essential aspects: refining dispersed document information into coherent and structured passages and preserving narrative flow by explicitly capturing and maintaining the temporal order among retrieved passages. We empirically demonstrate the effectiveness of ChronoRAG through experiments on the NarrativeQA and GutenQAdataset, showing substantial improvements in tasks requiring both factual identification and comprehension of complex sequential relationships, underscoring that reasoning over temporal order is crucial in resolving narrative QA.

cs.CL↗

arXiv2025

Investigating Language Preference of Multilingual RAG Systems

Multilingual Retrieval-Augmented Generation (mRAG) systems enhance language models by integrating external multilingual information to produce context-aware responses. However, mRAG systems struggle with retrieving relevant information due to linguistic variations between queries and documents, generating inconsistent responses when multilingual sources conflict. In this work, we systematically investigate language preferences in both retrieval and generation of mRAG through a series of experiments. Our analysis indicates that retrievers tend to prefer high-resource and query languages, yet this preference does not consistently improve generation performance. Moreover, we observe that generators prefer the query language or Latin scripts, leading to inconsistent outputs. To overcome these issues, we propose Dual Knowledge Multilingual RAG (DKM-RAG), a simple yet effective framework that fuses translated multilingual passages with complementary model knowledge. Empirical results demonstrate that DKM-RAG mitigates language preference in generation and enhances performance across diverse linguistic settings. Code is available at https://github.com/jeonghyunpark2002/LanguagePreference.git

cs.CL↗

arXiv2025

Conversational Query Reformulation with the Guidance of Retrieved Documents

Conversational search seeks to retrieve relevant passages for the given questions in conversational question answering. Conversational Query Reformulation (CQR) improves conversational search by refining the original queries into de-contextualized forms to resolve the issues in the original queries, such as omissions and coreferences. Previous CQR methods focus on imitating human written queries which may not always yield meaningful search results for the retriever. In this paper, we introduce GuideCQR, a framework that refines queries for CQR by leveraging key information from the initially retrieved documents. Specifically, GuideCQR extracts keywords and generates expected answers from the retrieved documents, then unifies them with the queries after filtering to add useful information that enhances the search process. Experimental results demonstrate that our proposed method achieves state-of-the-art performance across multiple datasets, outperforming previous CQR methods. Additionally, we show that GuideCQR can get additional performance gains in conversational search using various types of queries, even for queries written by humans.

cs.CL↗

arXiv2025

Radio-Frequency Pseudo-Null Induced by Light in an Ion Trap

In a linear radio-frequency (rf) ion trap, the rf null is the point of zero electric field in the dynamic trapping potential where the ion motion is approximately harmonic. When displaced from the rf null, the ion is superimposed by fast oscillations known as micromotion, which can be probed through motion-sensitive light-atom interactions. In this work, we report on the emergence of the rf pseudo-null, a locus of points where the ion responds to light as if it were at the true rf null, despite being displaced from it. The phenomenon is fully explained by accounting for the general two-dimensional structure of micromotion and is experimentally verified under various potential configurations, with observations in great agreement with numerical simulations. The rf pseudo-null manifests as a line in a two-dimensional parameter space, determined by the geometry of the incident light and its overlap with the motional structure of the ion. The true rf null occurs uniquely at the concurrent point of the pseudo-null lines induced by different light sources.

physics.atom-ph↗

arXiv2025

Efficient Quantum Frequency Conversion of Ultra-Violet Single Photons from a Trapped Ytterbium Ion

Ion trap system is a leading candidate for quantum network privileged by its long coherence time, high-fidelity gate operations, and the ion-photon entanglement that generates an ideal pair of a stationary memory qubit and a flying communication qubit. Rapid developments in nonlinear quantum frequency conversion techniques have enhanced the potential for constructing a trapped ion quantum network via optical fiber connections. The generation of long-distance entanglement has been demonstrated with ions such as Ca$^{+}$ and Ba$^{+}$, which emit photons in visible or near-infrared range naturally. On the other hand, as the qubit-native photons reside in ultra-violet (UV) spectrum, the Yb$^{+}$ ion has not been considered as a strong competitor for telecommunication qubits despite extensive research on it. Here, we demonstrate an efficient difference-frequency conversion of UV photons, emitted from a trapped Yb$^{+}$ ion, into a visible range. We provide experimental evidence that confirms the converted photons are radiated from the Yb$^{+}$ ion. Our results provide a crucial step toward realizing a long-distance trapped ion quantum network based on Yb$^{+}$ ions through quantum frequency conversion.

quant-ph↗

arXiv2024

A silicon-based ion trap chip protected from semiconductor charging

Silicon-based ion trap chips can benefit from existing advanced fabrication technologies, such as multi-metal layer techniques for two-dimensional architectures and silicon photonics for the integration of on-chip optical components. However, the scalability of these technologies may be compromised by semiconductor charging, where photogenerated charge carriers produce electric potentials that disrupt ion motion. Inspired by recent studies on charge distribution mechanisms in semiconductors, we developed a silicon-based chip with gold coated on all exposed silicon surfaces. This modification significantly stabilized ion motion compared to a chip without such metallic shielding, a result that underscores the detrimental effects of exposed silicon. With the mitigation of background silicon-induced fields to negligible levels, quantum operations such as sideband cooling and two-ion entangling gates, which were previously infeasible with the unshielded chip, can now be implemented.

quant-ph↗