arXiv subjects

Sangjun Park

Publications and source records attributed to Sangjun Park.

18 recordsLinked to original sources

arXiv2026

Position: Hippocampal Explicit Memory Is the Cornerstone for AGI

Large Language Models (LLMs) have demonstrated remarkable capabilities across various tasks, raising expectations for Artificial General Intelligence (AGI). This position paper argues that integrating explicit memory is the cornerstone for advancing LLMs toward AGI. The key reason is that the underlying learning mechanism of LLMs is highly analogous to human implicit memory. However, higher-order cognitive functions necessary for AGI, such as long-term strategic planning, metacognition, and symbolic reasoning, heavily rely on hippocampal explicit memory and cannot arise solely from implicit statistical learning. Drawing on findings from neuroscience, I advance this perspective and complement it with computational requirements for artificial explicit memory systems, hoping to foster further research and lay the groundwork for explicit memory integration.

cs.AI↗

arXiv2026

Diffusion Policy for Coordinated Control of a Nonholonomic Mobile Base and Dual Arms in Door Opening and Passing

Opening heavy, self closing doors, especially those that require pulling remains a long standing challenge in robotics. Humans naturally employ both arms in a dexterous manner, rotating the handle, widening the gap, holding the door, switching arms when needed, and moving through while maintaining clearance. To replicate such behaviors, a robot must perform a long sequence of motions spanning multiple stages and interactions with different parts of the door. Traditional approaches rely on state machines that transition between manually defined stages (e.g., pulling after the knob is rotated, passing after the gap is sufficiently wide). While intuitive, these methods lack robustness, as hand crafted trajectories fail to generalize to the diversity of real world conditions without extensive engineering effort. Recent advances in imitation learning offer a scalable alternative, yet no existing visual action model has demonstrated simultaneous coordination of a nonholonomic base and dual arms for the complete door opening and passing task. In this paper, we tackle this complex, highly constrained problem using a diffusion based visuomotor control policy. Our results demonstrate that a single end to end policy can be learned to execute long horizon tasks requiring tight coordination between manipulation and locomotion. The resulting policy not only achieves a high success rate in opening and traversing damped pull doors but also demonstrates strong robustness to external disturbances capabilities that are difficult to realize with traditional methods.

cs.RO↗

arXiv2026

Fine-Tuning Language Models to Know What They Know

Evaluating true metacognition in Large Language Models (LLMs) is difficult due to biases and heuristics. This paper presents a framework to measure and enhance LLM metacognition while controlling for these biases. A measurement method using the $d'_{\rm type2}$ metric is established to isolate metacognitive ability. The Evolution Strategy for Metacognitive Alignment (ESMA) is proposed, demonstrating robust generalization across unseen datasets, languages, and newly acquired knowledge. Finally, parameter analysis reveals that these improvements are driven by a sparse set of parameters, offering new pathways for targeted metacognitive optimization.

cs.NE↗

arXiv2025

Smart Contract-Enabled Procurement under Bounded Demand Variability: A Truncated Normal Approach

This study develops a strategic procurement framework integrating blockchain-based smart contracts with bounded demand variability modeled through a truncated normal distribution. While existing research emphasizes the technical feasibility of smart contracts, the operational and economic implications of adoption under moderate uncertainty remain underexplored. We propose a multi-supplier model in which a centralized retailer jointly determines the optimal smart contract adoption intensity and supplier allocation decisions. The formulation endogenizes adoption costs, supplier digital readiness, and inventory penalties to capture realistic trade-offs among efficiency, sustainability, and profitability. Analytical results establish concavity and provide closed-form comparative statics for adoption thresholds and procurement quantities. Extensive numerical experiments demonstrate that moderate demand variability supports partial adoption strategies, whereas excessive investment in digital infrastructure can reduce overall profitability. Dynamic simulations further reveal how adaptive learning and declining implementation costs progressively enhance adoption intensity and supply chain performance. The findings provide theoretical and managerial insights for balancing digital transformation, resilience, and sustainability objectives in smart contract-enabled procurement.

q-fin.GN↗

arXiv2025

Pigeon-SL: Robust Split Learning Framework for Edge Intelligence under Malicious Clients

Recent advances in split learning (SL) have established it as a promising framework for privacy-preserving, communication-efficient distributed learning at the network edge. However, SL's sequential update process is vulnerable to even a single malicious client, which can significantly degrade model accuracy. To address this, we introduce Pigeon-SL, a novel scheme grounded in the pigeonhole principle that guarantees at least one entirely honest cluster among M clients, even when up to N of them are adversarial. In each global round, the access point partitions the clients into N+1 clusters, trains each cluster independently via vanilla SL, and evaluates their validation losses on a shared dataset. Only the cluster with the lowest loss advances, thereby isolating and discarding malicious updates. We further enhance training and communication efficiency with Pigeon-SL+, which repeats training on the selected cluster to match the update throughput of standard SL. We validate the robustness and effectiveness of our approach under three representative attack models -- label flipping, activation and gradient manipulation -- demonstrating significant improvements in accuracy and resilience over baseline SL methods in future intelligent wireless networks.

cs.LG↗

arXiv2025

CReSt: A Comprehensive Benchmark for Retrieval-Augmented Generation with Complex Reasoning over Structured Documents

Large Language Models (LLMs) have made substantial progress in recent years, yet evaluating their capabilities in practical Retrieval-Augmented Generation (RAG) scenarios remains challenging. In practical applications, LLMs must demonstrate complex reasoning, refuse to answer appropriately, provide precise citations, and effectively understand document layout. These capabilities are crucial for advanced task handling, uncertainty awareness, maintaining reliability, and structural understanding. While some of the prior works address these aspects individually, there is a need for a unified framework that evaluates them collectively in practical RAG scenarios. To address this, we present CReSt (A Comprehensive Benchmark for Retrieval-Augmented Generation with Complex Reasoning over Structured Documents), a benchmark designed to assess these key dimensions holistically. CReSt comprises 2,245 human-annotated examples in English and Korean, designed to capture practical RAG scenarios that require complex reasoning over structured documents. It also introduces a tailored evaluation methodology to comprehensively assess model performance in these critical areas. Our evaluation shows that even advanced LLMs struggle to perform consistently across these dimensions, underscoring key areas for improvement. We release CReSt to support further research and the development of more robust RAG systems. The dataset and code are available at: https://github.com/UpstageAI/CReSt.

cs.CL↗

arXiv2025

Federated Learning Meets Fluid Antenna: Towards Robust and Scalable Edge Intelligence

Federated learning (FL) is an emerging machine learning paradigm with immense potential to support advanced services and applications in future industries. However, when deployed over wireless communication systems, FL suffers from significant communication overhead, which can be alleviated by integrating over-the-air computation (AirComp). Despite its advantages, AirComp introduces learning inaccuracies due to the inherent randomness of wireless channels, which can degrade overall learning performance. To address this issue, this paper explores the integration of fluid antenna systems (FAS) into AirComp-based FL to enhance system robustness and efficiency. Fluid antennas offer dynamic spatial diversity by adaptively selecting antenna ports, thereby mitigating channel variations and improving signal aggregation. Specifically, we propose an antenna selection rule for fluid-antenna-equipped devices that optimally enhances learning robustness or training performance. Building on this, we develop a learning algorithm and provide a theoretical convergence analysis. The simulation results validate the effectiveness of fluid antennas in improving FL performance, demonstrating their potential as a key enabler for wireless AI applications.

eess.SP↗

arXiv2024

MentalAgora: A Gateway to Advanced Personalized Care in Mental Health through Multi-Agent Debating and Attribute Control

As mental health issues globally escalate, there is a tremendous need for advanced digital support systems. We introduce MentalAgora, a novel framework employing large language models enhanced by interaction between multiple agents for tailored mental health support. This framework operates through three stages: strategic debating, tailored counselor creation, and response generation, enabling the dynamic customization of responses based on individual user preferences and therapeutic needs. We conduct experiments utilizing a high-quality evaluation dataset TherapyTalk crafted with mental health professionals, shwoing that MentalAgora generates expert-aligned and user preference-enhanced responses. Our evaluations, including experiments and user studies, demonstrate that MentalAgora aligns with professional standards and effectively meets user preferences, setting a new benchmark for digital mental health interventions.

cs.CL↗

arXiv2023

Memoria: Resolving Fateful Forgetting Problem through Human-Inspired Memory Architecture

Making neural networks remember over the long term has been a longstanding issue. Although several external memory techniques have been introduced, most focus on retaining recent information in the short term. Regardless of its importance, information tends to be fatefully forgotten over time. We present Memoria, a memory system for artificial neural networks, drawing inspiration from humans and applying various neuroscientific and psychological theories. The experimental results prove the effectiveness of Memoria in the diverse tasks of sorting, language modeling, and classification, surpassing conventional techniques. Engram analysis reveals that Memoria exhibits the primacy, recency, and temporal contiguity effects which are characteristics of human memory.

cs.LG↗

arXiv2022

Into-TTS : Intonation Template Based Prosody Control System

Intonations play an important role in delivering the intention of a speaker. However, current end-to-end TTS systems often fail to model proper intonations. To alleviate this problem, we propose a novel, intuitive method to synthesize speech in different intonations using predefined intonation templates. Prior to TTS model training, speech data are grouped into intonation templates in an unsupervised manner. Two proposed modules are added to the end-to-end TTS framework: an intonation predictor and an intonation encoder. The intonation predictor recommends a suitable intonation template to the given text. The intonation encoder, attached to the text encoder output, synthesizes speech abiding the requested intonation template. Main contributions of our paper are: (a) an easy-to-use intonation control system covering a wide range of users; (b) better performance in wrapping speech in a requested intonation with improved objective and subjective evaluation; and (c) incorporating a pre-trained language model for intonation modelling. Audio samples are available at https://srtts.github.io/IntoTTS.

eess.AS↗

arXiv2022

Bunched LPCNet2: Efficient Neural Vocoders Covering Devices from Cloud to Edge

Text-to-Speech (TTS) services that run on edge devices have many advantages compared to cloud TTS, e.g., latency and privacy issues. However, neural vocoders with a low complexity and small model footprint inevitably generate annoying sounds. This study proposes a Bunched LPCNet2, an improved LPCNet architecture that provides highly efficient performance in high-quality for cloud servers and in a low-complexity for low-resource edge devices. Single logistic distribution achieves computational efficiency, and insightful tricks reduce the model footprint while maintaining speech quality. A DualRate architecture, which generates a lower sampling rate from a prosody model, is also proposed to reduce maintenance costs. The experiments demonstrate that Bunched LPCNet2 generates satisfactory speech quality with a model footprint of 1.1MB while operating faster than real-time on a RPi 3B. Our audio samples are available at https://srtts.github.io/bunchedLPCNet2.

eess.AS↗

arXiv2021

Efficient valley polarization of charged excitons and resident carriers in MoS2 monolayers by optical pumping

We investigate with polarized microphotoluminescence the optical pumping of the valley degree of freedom in charge-tunable MoS2 monolayers encapsulated with hexagonal boron nitride at cryogenic temperatures. We report a large steady state valley polarization of the different excitonic complexes following circularly-polarized laser excitation 25 meV above the neutral exciton transition. For the first time in this material we reveal efficient valley pumping of positively-charged trions, which were so far elusive in non-encapsulated monolayers due to defect and laser-induced large electron doping. We find that negatively-charged trions present a polarization of 70 % which is unusually large for non-resonant excitation. We attribute this large valley polarization to the particular band structure of MoS2, where an optically dark exciton ground state coexists with a bright conduction band ordering in the single-particle picture, leading to a supression of the valley relaxation for negatively-charged trions. In addition, we demonstrate that circular excitation induces a dynamical polarization of resident electrons and holes, as recently shown in tungsten-based monolayers. This manifest itself as a variation in the intensity of different excitonic complexes under circular and linear excitation.

cond-mat.mes-hall↗

arXiv2021

Imaging Seebeck drift of excitons and trions in MoSe2 monolayers

Hyperspectral imaging at cryogenic temperatures is used to investigate exciton and trion propagation in MoSe$_2$ monolayers encapsulated with hexagonal boron nitride (hBN). Under a tightly focused, continuous-wave laser excitation, the spatial distribution of neutral excitons and charged trions strongly differ at high excitation densities. Remarkably, in this regime the trion distribution develops a halo shape, similar to that previously observed in WS2 monolayers at room temperature and under pulsed excitation. In contrast, the exciton distribution only presents a moderate broadening without the appereance of a halo. Spatially and spectrally resolved luminescence spectra reveal the buildup of a significant temperature gradient at high excitation power, that is attributed to the energy relaxation of photoinduced hot carriers. We show, via a numerical resolution of the transport equations for excitons and trions, that the halo can be interpreted as thermal drift of trions due to a Seebeck term in the particle current. The model shows that the difference between trion and exciton profiles is simply understood in terms of the very different lifetimes of these two quasiparticles.

cond-mat.mtrl-sci↗

arXiv2021

Spin/Valley pumping of resident electrons in WSe2 and WS2 monolayers

Monolayers of transition metal dichalcogenides are ideal materials to control both spin and valley degrees of freedom either electrically or optically. Nevertheless, optical excitation mostly generates excitons species with inherently short lifetime and spin/valley relaxation time. Here we demonstrate a very efficient spin/valley optical pumping of resident electrons in n-doped WSe2 and WS2 monolayers. We observe that, using a continuous wave laser and appropriate doping and excitation densities, negative trion doublet lines exhibit circular polarization of opposite sign and the photoluminescence intensity of the triplet trion is more than four times larger with circular excitation than with linear excitation. We interpret our results as a consequence of a large dynamic polarization of resident electrons using circular light.

cond-mat.mes-hall↗

arXiv2021

Ethereum ECCPoW

The error-correction code based proof-of-work (ECCPoW) algorithm is based on a low-density parity-check (LDPC) code. The ECCPoW is possible to impair ASIC with its time-varying capability of the parameters of LDPC code. Previous researches on the ECCPoW algorithm have presented its theory and implementation on Bitcoin. But they do not discuss how stable the block generation time is. A finite mean block generation time (BGT) and none heavy-tail BGT distribution are the ones of the focus in this study. In the ECCPoW algorithm, BGT may show a long-tailed distribution due to time-varying cryptographic puzzles. Thus, it is of interest to see if the BGT distribution is not heavy-tailed and if it shows a finite mean. If the distribution is heavy-tailed, then confirmation of a transaction cannot be guaranteed. We present implementation, simulation, and validation of ECCPoW Ethereum. In implementation, we explain how the ECCPoW algorithm is integrated into Ethereum 1.0 as a new consensus algorithm. In the simulation, we perform a multinode simulation to show that the ECCPoW Ethereum works well with automatic difficulty change. In the validation, we present the statistical results of the two-sample Anderson-Darling test to show that the distribution of BGT satisfies the necessary condition of the exponential distribution. Our implementation is downloadable at https://github.com/cryptoecc/ETH-ECC.

cs.CR↗

arXiv2020

Bunched LPCNet : Vocoder for Low-cost Neural Text-To-Speech Systems

LPCNet is an efficient vocoder that combines linear prediction and deep neural network modules to keep the computational complexity low. In this work, we present two techniques to further reduce it's complexity, aiming for a low-cost LPCNet vocoder-based neural Text-to-Speech (TTS) System. These techniques are: 1) Sample-bunching, which allows LPCNet to generate more than one audio sample per inference; and 2) Bit-bunching, which reduces the computations in the final layer of LPCNet. With the proposed bunching techniques, LPCNet, in conjunction with a Deep Convolutional TTS (DCTTS) acoustic model, shows a 2.19x improvement over the baseline run-time when running on a mobile device, with a less than 0.1 decrease in TTS mean opinion score (MOS).

eess.AS↗

arXiv2020

Time-Variant Proof-of-Work Using Error-Correction Codes

The protocol for cryptocurrencies can be divided into three parts, namely consensus, wallet, and networking overlay. The aim of the consensus part is to bring trustless rational peer-to-peer nodes to an agreement to the current status of the blockchain. The status must be updated through valid transactions. A proof-of-work (PoW) based consensus mechanism has been proven to be secure and robust owing to its simple rule and has served as a firm foundation for cryptocurrencies such as Bitcoin and Ethereum. Specialized mining devices have emerged, as rational miners aim to maximize profit, and caused two problems: i) the re-centralization of a mining market and ii) the huge energy spending in mining. In this paper, we aim to propose a new PoW called Error-Correction Codes PoW (ECCPoW) where the error-correction codes and their decoder can be utilized for PoW. In ECCPoW, puzzles can be intentionally generated to vary from block to block, leading to a time-variant puzzle generation mechanism. This mechanism is useful in repressing the emergence of the specialized mining devices. It can serve as a solution to the two problems of recentralization and energy spending.

cs.CR↗

arXiv2016

An Information Theoretic Study for Noisy Compressed Sensing With Joint Sparsity Model-2

In this paper, we study a support set reconstruction problem in which the signals of interest are jointly sparse with a common support set, and sampled by joint sparsity model-2 (JSM-2) in the presence of noise. Using mathematical tools, we develop upper and lower bounds on the failure probability of support set reconstruction in terms of the sparsity, the ambient dimension, the minimum signal to noise ratio, the number of measurement vectors and the number of measurements. These bounds can be used to provide a guideline to determine the system parameters in various applications of compressed sensing with noisy JSM-2. Based on the bounds, we develop necessary and sufficient conditions for reliable support set reconstruction. We interpret these conditions to give theoretical explanations about the benefits enabled by joint sparsity structure in noisy JSM-2. We compare our sufficient condition with the existing result of noisy multiple measurement vectors model (MMV). As a result, we show that noisy JSM-2 may require less number of measurements than noisy MMV for reliable support set reconstruction.

cs.IT↗