arXiv subjects

Suhyeon Lee

Publications and source records attributed to Suhyeon Lee.

At least 19 recordsLinked to original sources

arXiv2026

How You Ask Shapes What You Get: A Theory-Seeded Measurement of Articulation in Advice-Seeking LLM Conversations

Users articulate the same advice-seeking request in different ways: some specify detailed constraints, others gesture at a vague need. Prior work treats this variation as noise to be averaged away; we instead treat it as a stable, measurable structure in the input distribution. We ask whether articulation (how people ask) forms latent dimensions separable from topic (what they ask about), and whether it is associated with how language models respond. We extract interpretable features from 16,447 advice-seeking prompts pooled from public chat corpora (WildChat, LMSYS, and ShareChat) and recover a small set of latent articulation factors that replicate across train/test splits and across corpora. Because this structure is largely separable from topic, the populations it defines cut across topics and stay invisible to topic- or task-based evaluation. The factors define a handful of recurring articulation styles, one of which stands out: a long-form but information-poor style, roughly one in six prompts in the largest corpus, where models return shorter, vaguer answers and do not ask for clarification even though under-specification is exactly the condition that warrants it. The contrast holds within every topic group and length quintile, and is not under-specification alone -- a second, equally under-specified style does draw clarifying questions. Two independent human annotators reproduce this contrast. We argue that benchmarks should stratify on articulation, and we offer the extracted structure as a measurement instrument for doing so.

cs.CL↗

arXiv2026

LLM Pedagogical Behavior in AI Tutoring Interactions

Students increasingly use LLMs as tutors for coursework and problem solving. Little is known about the level of assistance LLMs provide when students use them as tutors in authentic learning interactions. This matters because tutoring responses can differ substantially in how directly they help students complete a task. We operationalize this dimension as scaffolding level and develop a five-level scale, validated against human annotations, that characterizes responses according to the degree of direct assistance they provide. We apply the scale to 14,637 LLM responses from 203 students in a university AI course. Responses are overwhelmingly concentrated at high levels of assistance, with more than 95% classified as either Explaining or Solving. Scaffolding level is systematically associated with students' subsequent conversational behavior, but provides little additional predictive information about performance on three subsequent exams beyond prior achievement and dialogue behavior. These findings provide an empirical baseline for LLM assistance in tutoring interactions and a measurement framework for evaluating how alternative tutoring designs change that assistance.

cs.CL↗

arXiv2026

Inside Qubic's Selfish Mining Campaign on Monero: Evidence, Tactics, and Limits

In 2025, the blockchain network Qubic conducted a mining campaign on Monero, a privacy-focused cryptocurrency. It publicly presented the campaign as a ``51\% takeover'' and a demonstration of selfish mining. The episode provides a rare opportunity to test such claims in a privacy-preserving proof-of-work network, where pool attribution and private release decisions are difficult to observe. We combine Monero node measurements, Qubic pool observations, community-shared artifacts, and disclosed view keys to attribute blocks and identify candidate withholding periods. Rather than sustained majority control, we observe elevated orphaning and deeper reorganizations. Qubic did not follow a single optimized selfish mining strategy, instead varying between standard selfish mining and more defensive block-release decisions. Across the campaign, it gained no reward advantage over honest mining. Community monitoring and Qubic's countermeasures reveal how real-world mining campaigns evolve in response to an active ecosystem. Overall, the campaign disrupted Monero without exhibiting a stable profit-maximizing strategy, leaving open disruptive incentives beyond direct mining rewards.

cs.CR↗

arXiv2026

InverseCrafter: Efficient Video ReCapture as a Latent Domain Inverse Problem

Recent approaches in controllable novel view video generation often rely on fine-tuning pre-trained Video Diffusion Models (VDMs). This dominant paradigm is computationally expensive and frequently suffers from catastrophic forgetting of the model's original generative priors. To address this challenge, here we propose InverseCrafter, a VDM training-free framework that reformulates novel view video generation as an inpainting-based inverse problem in the latent space, eliminating the need for any annotated 4D training data. The core of our method is to establish operator equivalence by employing a lightweight latent mask encoder to define a latent-domain masking operation via a continuous, multi-channel representation. This principled representation faithfully models the forward process in the latent domain, enabling efficient, backpropagation-free solvers while bypassing the costly bottleneck of repeated VAE operations. InverseCrafter achieves high-fidelity, spatio-temporally coherent novel view synthesis with near-zero additional inference overhead and excels at general-purpose video inpainting and editing by fully preserving the pre-trained VDM's generative capabilities.

cs.CV↗

arXiv2026

ViewSplat: View-Adaptive 3D Gaussian Splatting for Feed-Forward Synthesis

We present ViewSplat, a view-adaptive 3D Gaussian splatting network for novel view synthesis from unposed images. While recent feed-forward 3D Gaussian splatting has significantly accelerated 3D scene reconstruction by bypassing per-scene optimization, a fundamental fidelity gap remains. We attribute this gap to the limited capacity of single-step feed-forward networks to regress static Gaussian primitives that satisfy all viewpoints. To address this limitation, we shift the paradigm from static primitive regression to view-adaptive splatting. Instead of a rigid Gaussian representation, our pipeline learns a view-adaptive latent representation. Specifically, ViewSplat initially predicts base Gaussian primitives alongside the weights of scene-conditioned View MLPs. During rendering, these MLPs take target-view coordinates as input and predict view-dependent residual updates for each Gaussian attribute (i.e., 3D position, scale, rotation, opacity, and color). This mechanism, which we term view-adaptive splatting, allows each primitive to rectify initial estimation errors, effectively capturing high-fidelity appearances. Extensive experiments demonstrate that ViewSplat achieves state-of-the-art fidelity while maintaining fast inference and real-time rendering; our large backbone variant runs at 15 FPS during inference and 90 FPS during rendering. Our project page is available at https://cvlab-uos.github.io/ViewSplat.

cs.CV↗

arXiv2026

PromptLoop: Plug-and-Play Prompt Refinement via Latent Feedback for Diffusion Model Alignment

Despite recent progress, reinforcement learning (RL)-based fine-tuning of diffusion models often struggles with generalization, composability, and robustness against reward hacking. Recent studies have explored prompt refinement as a modular alternative, but most adopt a feed-forward approach that applies a single refined prompt throughout the entire sampling trajectory, thereby failing to fully leverage the sequential nature of reinforcement learning. To address this, we introduce PromptLoop, a plug-and-play RL framework that incorporates latent feedback into step-wise prompt refinement. Rather than modifying diffusion model weights, a multimodal large language model (MLLM) is trained with RL to iteratively update prompts based on intermediate latent states of diffusion models. This design achieves a structural analogy to the Diffusion RL approach, while retaining the flexibility and generality of prompt-based alignment. Extensive experiments across diverse reward functions and diffusion backbones demonstrate that PromptLoop (i) achieves effective reward optimization, (ii) generalizes seamlessly to unseen models, (iii) composes orthogonally with existing alignment methods, and (iv) mitigates over-optimization and reward hacking while introducing only a practically negligible inference overhead.

cs.LG↗

arXiv2026

(Im)possibility of Incentive Design for Challenge-based Blockchain Protocols

Blockchains offer a decentralized and secure execution environment strong enough to host cryptocurrencies, but the state-replication model makes on-chain computation expensive. To avoid heavy on-chain workloads, systems like Truebit and optimistic rollups use challenge-based protocols, performing computations off-chain and invoking the chain only when challenged. This keeps normal-case costs low and, if at least one honest challenger exists, can catch fraud. What has been less clear is whether honest challengers are actually incentivized and a dishonest proposer is properly damaged under the worst case environment. We build a model with a colluding minority, heterogeneous costs, and three ordering modes. We then ask whether two goals can be met together: honest non-loss and fraud deterrence. Our results are clear: in single-winner designs, the incentive design is impossible or limited in scale. By contrast, in multi-winner designs, we obtain simple, explicit conditions under which both goals hold.

cs.GT↗

arXiv2025

Looking for Attention: Randomized Attention Test Design for Validator Monitoring in Optimistic Rollups

Optimistic Rollups (ORUs) significantly enhance blockchain scalability but inherently suffer from the verifier's dilemma, particularly concerning validator attentiveness. Current systems lack mechanisms to proactively ensure validators are diligently monitoring L2 state transitions, creating a vulnerability where fraudulent states could be finalized. This paper introduces the Randomized Attention Test (RAT), a novel L1-based protocol designed to probabilistically challenge validators in ORUs, thereby verifying their liveness and computational readiness. Our game-theoretic analysis demonstrates that an Ideal Security Equilibrium, where all validators are attentive and proposers are honest, can be achieved with RAT. Notably, this equilibrium is attainable and stable with relatively low economic penalties (under \$1000) for non-responsive validators, a low attention test frequency (under 1\% per epoch), and a minimal operation overhead (monthly under \$30) with 10 validators. RAT thus provides a pivotal, practical mechanism to enforce validator diligence, fortifying the overall security and integrity of ORU systems with minimizing additional costs.

cs.CR↗

arXiv2025

Commit-Reveal$^2$: Securing Randomness Beacons with Randomized Reveal Order in Smart Contracts

Simple commit-reveal beacons are vulnerable to last-revealer strategies, and existing descriptions often leave accountability and recovery mechanisms unspecified for practical deployments. We present Commit-Reveal$^2$, a layered design for blockchain deployments that cryptographically randomizes the final reveal order, together with a concrete accountability and fallback mechanism that we implement as smart-contract logic. The protocol is architected as a hybrid system, where routine coordination runs off chain for efficiency and the blockchain acts as the trust anchor for commitments and the final arbiter for disputes. Our implementation covers leader coordination, on-chain verification, slashing for non-cooperation, and an explicit on-chain recovery path that maintains progress when off-chain coordination fails. We formally define two security goals for distributed randomness beacons, unpredictability and bit-wise bias resistance, and we show that Commit-Reveal$^2$ meets these notions under standard hash assumptions in the random-oracle model. In measurements with small to moderate operator sets, the hybrid design reduces on-chain gas by more than 80% compared to a fully on-chain baseline. We release a publicly verifiable prototype and evaluation artifacts to support replication and adoption in blockchain applications.

cs.CR↗

arXiv2025

Reangle-A-Video: 4D Video Generation as Video-to-Video Translation

We introduce Reangle-A-Video, a unified framework for generating synchronized multi-view videos from a single input video. Unlike mainstream approaches that train multi-view video diffusion models on large-scale 4D datasets, our method reframes the multi-view video generation task as video-to-videos translation, leveraging publicly available image and video diffusion priors. In essence, Reangle-A-Video operates in two stages. (1) Multi-View Motion Learning: An image-to-video diffusion transformer is synchronously fine-tuned in a self-supervised manner to distill view-invariant motion from a set of warped videos. (2) Multi-View Consistent Image-to-Images Translation: The first frame of the input video is warped and inpainted into various camera perspectives under an inference-time cross-view consistency guidance using DUSt3R, generating multi-view consistent starting images. Extensive experiments on static view transport and dynamic camera control show that Reangle-A-Video surpasses existing methods, establishing a new solution for multi-view video generation. We will publicly release our code and data. Project page: https://hyeonho99.github.io/reangle-a-video/

cs.CV↗

arXiv2025

A Comment on "e-PoS: Making PoS Decentralized and Fair"

Proof-of-Stake (PoS) is a prominent Sybil control mechanism for blockchain-based systems. In "e-PoS: Making PoS Decentralized and Fair," Saad et al. (TPDS'21) introduced a new Proof-of-Stake protocol, e-PoS, to enhance PoS applications' decentralization and fairness. In this comment paper, we address a misunderstanding in the work of Saad et al. The conventional Proof-of-Stake model that causes the fairness problem does not align with the general concept of Proof-of-Stake nor the Proof-of-Stake cryptocurrencies mentioned in their paper.

cs.CR↗

arXiv2025

Hollow Victory: How Malicious Proposers Exploit Validator Incentives in Optimistic Rollup Dispute Games

Blockchain systems, such as Ethereum, are increasingly adopting layer-2 scaling solutions to improve transaction throughput and reduce fees. One popular layer-2 approach is the Optimistic Rollup, which relies on a mechanism known as a dispute game for block proposals. In these systems, validators can challenge blocks that they believe contain errors, and a successful challenge results in the transfer of a portion of the proposer's deposit as a reward. In this paper, we reveal a structural vulnerability in the mechanism: validators may not be awarded a proper profit despite winning a dispute challenge. We develop a formal game-theoretic model of the dispute game and analyze several scenarios, including cases where the proposer controls some validators and cases where a secondary auction mechanism is deployed to induce additional participation. Our analysis demonstrates that under current designs, the competitive pressure from validators may be insufficient to deter malicious behavior. We find that increased validator competition, paradoxically driven by higher rewards or participation, can allow a malicious proposer to significantly lower their net loss by capturing value through mechanisms like auctions. To address this, we propose countermeasures such as an escrowed reward mechanism and a commit-reveal protocol. Our findings provide critical insights into enhancing the economic security of layer-2 scaling solutions in blockchain networks.

cs.GT↗

arXiv2025

Single-Step Bidirectional Unpaired Image Translation Using Implicit Bridge Consistency Distillation

Unpaired image-to-image translation has seen significant progress since the introduction of CycleGAN. However, methods based on diffusion models or Schrödinger bridges have yet to be widely adopted in real-world applications due to their iterative sampling nature. To address this challenge, we propose a novel framework, Implicit Bridge Consistency Distillation (IBCD), which enables single-step bidirectional unpaired translation without using adversarial loss. IBCD extends consistency distillation by using a diffusion implicit bridge model that connects PF-ODE trajectories between distributions. Additionally, we introduce two key improvements: 1) distribution matching for consistency distillation and 2) adaptive weighting method based on distillation difficulty. Experimental results demonstrate that IBCD achieves state-of-the-art performance on benchmark datasets in a single generation step. Project page available at https://hyn2028.github.io/project_page/IBCD/index.html

cs.CV↗

arXiv2025

Implementation Study of Cost-Effective Verification for Pietrzak's Verifiable Delay Function in Ethereum Smart Contracts

Verifiable Delay Function (VDF) is a cryptographic concept that ensures a minimum delay before output through sequential processing, which is resistant to parallel computing. One of the significant VDF protocols academically reviewed is the VDF protocol proposed by Pietrzak. However, for the blockchain environment, the Pietrzak VDF has drawbacks including long proof size and recursive protocol computation. In this paper, we present an implementation study of Pietrzak VDF verification on Ethereum Virtual Machine (EVM). We found that the discussion in the Pietrzak's original paper can help a clear optimization in EVM where the costs of computation are predefined as the specific amounts of gas. In our results, the cost of VDF verification can be reduced from 4M to 2M gas, and the proof length can be generated under 8 KB with the 2048-bit RSA key length, which is much smaller than the previous expectation.

cs.CR↗

arXiv2024

180 Days After EIP-4844: Will Blob Sharing Solve Dilemma for Small Rollups?

The introduction of blobs through EIP-4844 has significantly reduced the Data Availability (DA) costs for rollups on Ethereum. However, due to the fixed size of blobs at 128 KB, rollups with low data throughput face a dilemma: they either use blobs inefficiently or decrease the frequency of DA submissions. Blob sharing, where multiple rollups share a single blob, has been proposed as a solution to this problem. This paper examines the effectiveness of blob sharing based on real-world data collected approximately six months after the implementation of EIP-4844. By simulating cost changes using a simple blob sharing format, we demonstrate that blob sharing can substantially improve the costs and DA service quality for small rollups, effectively resolving their dilemma. Notably, we observed cost reductions in USD exceeding 85% for most of the rollups when they cooperate, attributable to the smoothing effect of the blob base fee achieved through blob sharing.

cs.CR↗

arXiv2024

LLM-CXR: Instruction-Finetuned LLM for CXR Image Understanding and Generation

Following the impressive development of LLMs, vision-language alignment in LLMs is actively being researched to enable multimodal reasoning and visual IO. This direction of research is particularly relevant to medical imaging because medical image analysis and generation consist of reasoning based on a combination of visual features and prior knowledge. Many recent works have focused on training adapter networks that serve as an information bridge between image processing networks and LLMs; but presumably, in order to achieve maximum reasoning potential of LLMs on visual information as well, visual and language features should be allowed to interact more freely. This is especially important in the medical domain because understanding and generating medical images such as chest X-rays (CXR) require not only accurate visual and language-based reasoning but also a more intimate mapping between the two modalities. Thus, taking inspiration from previous work on the transformer and VQ-GAN combination for bidirectional image and text generation, we build upon this approach and develop a method for instruction-tuning an LLM pre-trained only on text to gain vision-language capabilities for medical images. Specifically, we leverage a pretrained LLM's existing question-answering and instruction-following abilities to teach it to understand visual inputs by instructing it to answer questions about image inputs and, symmetrically, output both text and image responses appropriate to a given query by tuning the LLM with diverse tasks that encompass image-based text-generation and text-based image-generation. We show that our model, LLM-CXR, trained in this approach shows better image-text alignment in both CXR understanding and generation tasks while being smaller in size compared to previously developed models that perform a narrower range of tasks. The code is at https://github.com/hyn2028/llm-cxr.

cs.CV↗

arXiv2024

A Tip for IOTA Privacy: IOTA Light Node Deanonymization via Tip Selection

IOTA is a distributed ledger technology that uses a Directed Acyclic Graph (DAG) structure called the Tangle. It is known for its efficiency and is widely used in the Internet of Things (IoT) environment. Tangle can be configured by utilizing the tip selection process. Due to performance issues with light nodes, full nodes are being asked to perform the tip selections of light nodes. However, in this paper, we demonstrate that tip selection can be exploited to compromise users' privacy. An adversary full node can associate a transaction with the identity of a light node by comparing the light node's request with its ledger. We show that these types of attacks are not only viable in the current IOTA environment but also in IOTA 2.0 and the privacy improvement being studied. We also provide solutions to mitigate these attacks and propose ways to enhance anonymity in the IOTA network while maintaining efficiency and scalability.

cs.CR↗

arXiv2024

Don't Play Favorites: Minority Guidance for Diffusion Models

We explore the problem of generating minority samples using diffusion models. The minority samples are instances that lie on low-density regions of a data manifold. Generating a sufficient number of such minority instances is important, since they often contain some unique attributes of the data. However, the conventional generation process of the diffusion models mostly yields majority samples (that lie on high-density regions of the manifold) due to their high likelihoods, making themselves ineffective and time-consuming for the minority generating task. In this work, we present a novel framework that can make the generation process of the diffusion models focus on the minority samples. We first highlight that Tweedie's denoising formula yields favorable results for majority samples. The observation motivates us to introduce a metric that describes the uniqueness of a given sample. To address the inherent preference of the diffusion models w.r.t. the majority samples, we further develop minority guidance, a sampling technique that can guide the generation process toward regions with desired likelihood levels. Experiments on benchmark real datasets demonstrate that our minority guidance can greatly improve the capability of generating high-quality minority samples over existing generative samplers. We showcase that the performance benefit of our framework persists even in demanding real-world scenarios such as medical imaging, further underscoring the practical significance of our work. Code is available at https://github.com/soobin-um/minority-guidance.

cs.LG↗