arXiv subjects

Manami Nakagawa

Publications and source records attributed to Manami Nakagawa.

7 recordsLinked to original sources

arXiv2026

VArify: A Visual Analytics System for Verifying Knowledge Enhanced Large Language Model Responses in Food Science

Graph Retrieval-Augmented Generation (GraphRAG) enables Large Language Models (LLMs) to leverage structured, domain-specific knowledge graph databases for factually grounded responses. However, the retrieval of irrelevant or conflicting data can still result in erroneous responses. In knowledge-intensive and evidence-focused domains, human verification of the supporting evidence for an LLM response is still necessary. We conducted a formative pilot study to characterize the challenges of verifying complex, multi-layered data retrieved by GraphRAG systems. Based on these insights, we present VArify, a visual analytics system that leverages a file directory-inspired tree visualization to support simultaneous exploration of inter-group relationships and intra-group hierarchies within the retrieved evidence. We evaluate VArify through a user study with six food science experts and students. Our results indicate that the system effectively helps users distinguish between an LLM's internal parametric knowledge and external graph-sourced evidence. Furthermore, the visualization helped experts identify inaccuracies within the underlying knowledge graph itself, leading to more calibrated trust in the model's output. We conclude by discussing opportunities to leverage visualizations to further support verification regarding unknown unknowns, personalization, and limitations of knowledge graphs.

cs.HC↗

arXiv2025

LAPPI: Interactive Optimization with LLM-Assisted Preference-Based Problem Instantiation

Many real-world tasks, such as trip planning or meal planning, can be formulated as combinatorial optimization problems. However, using optimization solvers is difficult for end users because it requires problem instantiation: defining candidate items, assigning preference scores, and specifying constraints. We introduce LAPPI (LLM-Assisted Preference-based Problem Instantiation), an interactive approach that uses large language models (LLMs) to support users in this instantiation process. Through natural language conversations, the system helps users transform vague preferences into well-defined optimization problems. These instantiated problems are then passed to existing optimization solvers to generate solutions. In a user study on trip planning, our method successfully captured user preferences and generated feasible plans that outperformed both conventional and prompt-engineering approaches. We further demonstrate LAPPI's versatility by adapting it to an additional use case.

cs.HC↗

arXiv2025

Binding energy of $^{3}_Λ\rm{H}$ and $^{4}_Λ\rm{H}$ via image analyses of nuclear emulsions using deep-learning

Subatomic systems are pivotal for understanding fundamental baryonic interactions, as they provide direct access to quark-level degrees of freedom. In particular, introducing a strange quark adds "strangeness" as a new dimension, offering a powerful tool for exploring nuclear forces. The hypertriton, the lightest three-body hypernuclear system, provides an ideal testing ground for investigating baryonic interactions and quark behavior involving up, down, and strange quarks. However, experimental measurements of its lifetime and binding energy, key indicators of baryonic interactions, show significant deviations in results obtained from energetic collisions of heavy-ion beams. Identifying alternative pathways for precisely measuring the hypertriton's binding energy and lifetime is thus crucial for advancing experimental and theoretical nuclear physics. Here, we present an experimental study on the binding energies of $^3_Λ\mathrm{H}$ (hypertriton) and $^4_Λ\mathrm{H}$, performed through the analysis of photographic nuclear emulsions using modern techniques. By incorporating deep-learning methods, we uncovered systematic uncertainties in conventional nuclear emulsion analyses and established a refined calibration protocol for determining binding energies accurately. Our results are independent of those obtained from heavy-ion collision experiments, offering a complementary measurement and opening new avenues for investigating few-body hypernuclei interactions.

nucl-ex↗

arXiv2024

A novel machine learning method to detect double-$Λ$ hypernuclear events in nuclear emulsions

A novel method was developed to detect double-$Λ$ hypernuclear events in nuclear emulsions using machine learning techniques. The object detection model, the Mask R-CNN, was trained using images generated by Monte Carlo simulations, image processing, and image-style transformation based on generative adversarial networks. Despite being exclusively trained on $\prescript{6\ }{ΛΛ}{\rm{He}}$ events, the model achieved a detection efficiency of 93.8$\%$ for $\prescript{6\ }{ΛΛ}{\rm{He}}$ and 82.0$\%$ for $\prescript{5\ }{ΛΛ}{\rm{H}}$ events in the produced images. In addition, the model demonstrated its ability to detect the $\prescript{6\ }{ΛΛ}{\rm{He}}$ event named the Nagara event, which is the only uniquely identified double-$Λ$ hypernuclear event reported to date. It also exhibited a proper segmentation of the event topology. Furthermore, after analyzing 0.2$\%$ of the entire emulsion data from the J-PARC E07 experiment utilizing the developed approach, six new candidates for double-$Λ$ hypernuclear events were detected, suggesting that more than 2000 double-strangeness hypernuclear events were recorded in the entire dataset. This method is sufficiently effective for mining more latent double-$Λ$ hypernuclear events recorded in nuclear emulsion sheets by reducing the time required for manual visual inspection by a factor of five hundred.

hep-ex↗

arXiv2022

Investigation of the neutron imaging applications using fine-grained nuclear emulsion

Neutron imaging is a non-destructive inspection technique with a wide range of applications. One of the important aspects concerning neutron imaging is achieving micrometer-scale spatial resolution. Developing a neutron detector with a high resolution is a challenging task. Neutron detectors, based on fine-grained nuclear emulsion, may be suitable for high resolution neutron imaging applications. High track density is a necessary requirement to improve the quality of neutron imaging. However, the available track analysis methods are difficult to apply under high track density conditions. Simulated images were used to determine the required track density for neutron imaging. It was concluded that a track density of the order of $10^4$ tracks per 100 $\times$ 100 $μ$m$^2$ is sufficient to utilize neutron detectors for imaging applications. The contrast resolution was also investigated for the image data sets with various track densities and neutron transmission rates. Moreover, experiments were performed for neutron imaging of the gadolinium-based gratings with known geometries. The structure of gratings was successfully resolved. The calculated 1$σ$ 10-90 \% edge response, using the gray scale optical images of the grating slit with a periodic structure of 9 $μ$m, was 0.945 $\pm$ 0.004 $μ$m.

physics.ins-det↗

arXiv2022

Study on the reusability of fluorescent nuclear track detectors using optical bleaching

Fluorescent nuclear track detectors (FNTDs) based on Al${_2}$O${_3}$:C,Mg crystals are luminescent detectors that can be used for dosimetry and detection of charged particles and neutrons. These detectors can be utilised for imaging applications where a reasonably high track density, approximately of the order of 1 $\times$ $10^4$ tracks in an area of 100 $\times$ 100 $μ$m$^2$, is required. To investigate the reusability of FNTDs for imaging applications, we present an approach to perform optical bleaching under the required track density conditions. The reusability was assessed through seven irradiation-bleaching cycles. For the irradiation, the studied FNTD was exposed to alpha-particles from an $^{241}$Am radioactive source. The optical bleaching was performed by means of ultraviolet laser light with a wavelength of 355 nm. Three dedicated regions on a single FNTD with different accumulated track densities and bleaching conditions were investigated. After every irradiation-bleaching cycle, signal-to-noise ratio was calculated to evaluate FNTD performance. It is concluded that FNTDs can be reused at least seven times for applications where accumulation of a high track density is required.

physics.ins-det↗

arXiv2021

Novel method for producing very-neutron-rich hypernuclei via charge-exchange reactions with heavy ion projectiles

We propose a novel method for producing very-neutron-rich hypernuclei and corresponding resonance states by employing charge-exchange reactions via pp($^{12}$C, $^{12}$N $K^+$)n$Λ$ with single-charge-exchange and ppp($^{9}$Be, $^{9}$C $K^+$)nn$Λ$ with double-charge-exchange, both of which produce $ΛK^+$ in a target nucleus. The feasibility of producing very-neutron-rich hypernuclei using the proposed method was analysed by applying an ultra-relativistic quantum molecular dynamics model to a $^6$Li+$^{12}$C reaction at 2 $A$ GeV. The yields of very-neutron-rich hypernuclei, signal-to-background ratios, and background contributions were investigated. The proposed method is a powerful tool for studying very-neutron-rich hypernuclei and resonance states with a hyperon for experiments employing the Super-FRS facility at FAIR and HFRS facility at HIAF.

nucl-ex↗