Searcharxiv⌕ Search

arXiv subjects

Do-Hyung Kim

Publications and source records attributed to Do-Hyung Kim.

16 recordsLinked to original sources
arXiv2026

SemKV: Semantic Mixed-Precision KV Cache Quantization Guided by the Quality Cliff for Long-Context LLM Inference

The key-value (KV) cache is the dominant memory bottleneck of long-context large language model (LLM) inference, growing linearly with context length. We show that uniform KV quantization on a fractional-bit grid does not degrade gracefully: under a prespecified multi-seed statistical protocol, Llama-3.1-8B-Instruct with an affine quantizer is statistically indistinguishable from FP16 KV down to 2.322 code bits/value and collapses at 2.0 bits - a quality cliff in (2.0, 2.322] that reappears in generation-time quantization and multi-turn dialogue and transfers to Mistral-7B. The cliff reframes importance-aware mixed precision: above it, eight model-internal importance indicators are statistically interchangeable, so the benefit of mixing is grid interpolation, reaching average precisions uniform quantization cannot realize. SemKV preserves every token, ranks tokens by a model-internal score, and assigns two adjacent above-cliff precisions, achieving a measured 6.0x storage reduction with no statistically detectable quality difference from full KV (n=900, three seeds), and outperforming FP16 token pruning granted a 1.5x larger memory budget. Replacing the affine base with a distortion-optimized quantizer (TurboQuant-MSE) lowers the cliff in every protocol tested, raising the no-detectable-loss operating point to 7.9x. The recipe: measure the cliff for the target deployment setting, then interpolate above it.

cs.LG↗
arXiv2025

Symplectic Homology and 3-dimensional Besse Manifolds with vanishing first Chern class

In this paper, we will show that certain types of symplectic homology can be used as an invariant of 3-dimensional Besse manifolds, which are strict contact manifolds with periodic Reeb flow. For simplicity, we will assume our Besse structures to be a trivial plane bundle. To identify Besse manifolds with such a condition, we actually compute the first Chern class of each Besse structure and classify the Besse manifolds with vanishing first Chern class. We will also compute Robbin-Salamon indices of periodic Reeb orbits in Besse manifolds, and symplectic homology (of its filling) of certain cases. From its definition, Besse manifolds naturally become Seifert fibration and thus one can extract invariants such as the orbifold Euler characteristic and the Euler number of this Seifert fibration. These invariants will become important in our computation.

math.SG↗
arXiv2025

A Multi-tiered Human-in-the-loop Approach for Interactive School Mapping Using Earth Observation and Machine Learning

This paper presents a multi-tiered human-in-the-loop framework for interactive school mapping designed to improve the accuracy and completeness of educational facility records, particularly in developing regions where such data may be scarce and infrequently updated. The first tier involves a machine learning based analysis of population density, land cover, and existing infrastructure compared with known school locations. The first tier identifies potential gaps and "mislabelled" schools. In subsequent tiers, medium-resolution satellite imagery (Sentinel-2) is investigated to pinpoint regions with a high likelihood of school presence, followed by the application of very high-resolution (VHR) imagery and deep learning models to generate detailed candidate locations for schools within these prioritised areas. The medium-resolution approach was later removed due to insignificant improvements. The medium and VHR resolution models build upon global pre-trained steps to improve generalisation. A key component of the proposed approach is an interactive interface to allow human operators to iteratively review, validate, and refine the mapping results. Preliminary evaluations indicate that the multi-tiered strategy provides a scalable and cost-effective solution for educational infrastructure mapping to support planning and resource allocation.

cs.CV↗
arXiv2024

Predicting Internet Connectivity in Schools: A Feasibility Study Leveraging Multi-modal Data and Location Encoders in Low-Resource Settings

Internet connectivity in schools is critical to provide students with the digital literary skills necessary to compete in modern economies. In order for governments to effectively implement digital infrastructure development in schools, accurate internet connectivity information is required. However, traditional survey-based methods can exceed the financial and capacity limits of governments. Open-source Earth Observation (EO) datasets have unlocked our ability to observe and understand socio-economic conditions on Earth from space, and in combination with Machine Learning (ML), can provide the tools to circumvent costly ground-based survey methods to support infrastructure development. In this paper, we present our work on school internet connectivity prediction using EO and ML. We detail the creation of our multi-modal, freely-available satellite imagery and survey information dataset, leverage the latest geographically-aware location encoders, and introduce the first results of using the new European Space Agency phi-lab geographically-aware foundational model to predict internet connectivity in Botswana and Rwanda. We find that ML with EO and ground-based auxiliary data yields the best performance in both countries, for accuracy, F1 score, and False Positive rates, and highlight the challenges of internet connectivity prediction from space with a case study in Kigali, Rwanda. Our work showcases a practical approach to support data-driven digital infrastructure development in low-resource settings, leveraging freely available information, and provide cleaned and labelled datasets for future studies to the community through a unique collaboration between UNICEF and the European Space Agency phi-lab.

eess.IV↗
arXiv2024

Uncovering large inconsistencies between machine learning derived gridded settlement datasets

High-resolution human settlement maps provide detailed delineations of where people live and are vital for scientific and practical purposes, such as rapid disaster response, allocation of humanitarian resources, and international development. The increased availability of high-resolution satellite imagery, combined with powerful techniques from machine learning and artificial intelligence, has spurred the creation of a wealth of settlement datasets. However, the precise agreement and alignment between these datasets is not known. Here we quantify the overlap of high-resolution settlement map for 42 African countries developed by Google (Open Buildings), Meta (High Resolution Population Maps) and GRID3 (Geo-Referenced Infrastructure and Demographic Data for Development). Across all studied countries we find large disagreement between datasets on how much area is considered settled. We demonstrate that there are considerable geographic and socio-economic factors at play and build a machine learning model to predict for which areas datasets disagree. It it vital to understand the shortcomings of AI derived high-resolution settlement layers as international organizations, governments, and NGOs are already experimenting with incorporating these into programmatic work. As such, we anticipate our work to be a starting point for more critical and detailed analyses of AI derived datasets for humanitarian, planning, policy, and scientific purposes.

cs.SI↗
arXiv2021

Towards an Open Global Air Quality Monitoring Platform to Assess Children's Exposure to Air Pollutants in the Light of COVID-19 Lockdowns

This ongoing work attempts to understand and address the requirements of UNICEF, a leading organization working in children's welfare, where they aim to tackle the problem of air quality for children at a global level. We are motivated by the lack of a proper model to account for heavily fluctuating air quality levels across the world in the wake of the COVID-19 pandemic, leading to uncertainty among public health professionals on the exact levels of children's exposure to air pollutants. We create an initial model as per the agency's requirement to generate insights through a combination of virtual meetups and online presentations. Our research team comprised of UNICEF's researchers and a group of volunteer data scientists. The presentations were delivered to a number of scientists and domain experts from UNICEF and community champions working with open data. We highlight their feedback and possible avenues to develop this research further.

cs.HC↗
arXiv2016

On continuous causal isomorphisms

It is shown that continuous causal isomorphisms on two-dimensional Minkowski spacetime can be characterized by the invariance of wave equations.

math-ph↗
arXiv2015

Causal and conformal structures of globally hyperbolic spacetimes

The group of conformal diffeomorphisms and the group of causal automorphisms on two-dimensional globally hyperbolic spacetimes are clarified. It is shown that if spacetimes have non-compact Cauchy surfaces, then the groups are subgroups of that of two-dimensional Minkowski spacetime, and if spacetimes have compact Cauchy surfaces, then the groups are subgroups of that of two-dimensional Einstein's static universe.

math.DG↗
arXiv2015

An imbedding of spacetimes

It is shown that any two-dimensional spacetimes with compact Cauchy surfaces can be causally isomorphically imbedded into the two-dimensional Einstein's static universe. Also, it is shown that any two-dimensional globally hyperbolic spacetimes are conformally equivalent to a subset of the two-dimensional Einstein's static universe.

math-ph↗
arXiv2013

A note on conformal equivalences

A new characterization of conformal transformations is given. By use of this, the general form of conformal transformation on two-dimensional Minkowski space is given and its conformal structure is analyzed.

math.DG↗
arXiv2013

Relativity on two-dimensional spacetimes

Lorentz transformation on two-dimensional spacetime is obtained without assumption of linearity. To obtain this, we use the invariance of wave equations, which is recently proved to be equivalent to the causality preservation.

gr-qc↗
arXiv2008

A Note on Non-compact Cauchy surface

It is shown that if a space-time has non-compact Cauchy surface, then its topological, differentiable, and causal structure are completely determined by a class of compact subsets of its Cauchy surface. Since causal structure determines its topological, differentiable, and conformal structure of space-time, this gives a natural way to encode the corresponding structures into its Cauchy surface.

gr-qc↗