Kimin Lee
YOU?
Author Swipe
View article: Adversarial Reinforcement Learning Framework for ESP Cheater Simulation
Adversarial Reinforcement Learning Framework for ESP Cheater Simulation Open
Extra-Sensory Perception (ESP) cheats, which reveal hidden in-game information such as enemy locations, are difficult to detect because their effects are not directly observable in player behavior. The lack of observable evidence makes it …
View article: Learning to Generate Unit Test via Adversarial Reinforcement Learning
Learning to Generate Unit Test via Adversarial Reinforcement Learning Open
Unit testing is a core practice in programming, enabling systematic evaluation of programs produced by human developers or large language models (LLMs). Given the challenges in writing comprehensive unit tests, LLMs have been employed to a…
View article: Unintended Misalignment from Agentic Fine-Tuning: Risks and Mitigation
Unintended Misalignment from Agentic Fine-Tuning: Risks and Mitigation Open
Beyond simple text generation, Large Language Models (LLMs) have evolved into agentic systems capable of planning and interacting with external tools to solve complex tasks. This evolution involves fine-tuning LLMs on agent-specific tasks …
View article: Pulsed electric field pretreatment to enhance sodium chloride and moisture diffusion in radish tissues while maintaining microstructural integrity
Pulsed electric field pretreatment to enhance sodium chloride and moisture diffusion in radish tissues while maintaining microstructural integrity Open
This study examined the influence of pulsed electric field (PEF) pretreatment on sodium chloride (NaCl) and moisture mass transfer in radish (Raphanus sativus L.) tissues, aiming to improve salting efficiency while preserving microstructur…
View article: Enhancing Motion Dynamics of Image-to-Video Models via Adaptive Low-Pass Guidance
Enhancing Motion Dynamics of Image-to-Video Models via Adaptive Low-Pass Guidance Open
Recent text-to-video (T2V) models have demonstrated strong capabilities in producing high-quality, dynamic videos. To improve the visual controllability, recent works have considered fine-tuning pre-trained T2V models to support image-to-v…
View article: Prime the search: Using large language models for guiding geometric task and motion planning by warm-starting tree search
Prime the search: Using large language models for guiding geometric task and motion planning by warm-starting tree search Open
The problem of relocating a set of objects to designated areas amidst movable obstacles can be framed as a Geometric Task and Motion Planning ( g-tamp ), a subclass of task and motion planning problem (TAMP). Traditional approaches to g-ta…
View article: Revisiting Multi-Agent Debate as Test-Time Scaling: A Systematic Study of Conditional Effectiveness
Revisiting Multi-Agent Debate as Test-Time Scaling: A Systematic Study of Conditional Effectiveness Open
The remarkable growth in large language model (LLM) capabilities has spurred exploration into multi-agent systems, with debate frameworks emerging as a promising avenue for enhanced problem-solving. These multi-agent debate (MAD) approache…
View article: What Really Matters in Many-Shot Attacks? An Empirical Study of Long-Context Vulnerabilities in LLMs
What Really Matters in Many-Shot Attacks? An Empirical Study of Long-Context Vulnerabilities in LLMs Open
We investigate long-context vulnerabilities in Large Language Models (LLMs) through Many-Shot Jailbreaking (MSJ). Our experiments utilize context length of up to 128K tokens. Through comprehensive analysis with various many-shot attack set…
View article: DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models
DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models Open
Fine-tuning text-to-image diffusion models to maximize rewards has proven effective for enhancing model performance. However, reward fine-tuning methods often suffer from slow convergence due to online sample generation. Therefore, obtaini…
View article: Silent Branding Attack: Trigger-free Data Poisoning Attack on Text-to-Image Diffusion Models
Silent Branding Attack: Trigger-free Data Poisoning Attack on Text-to-Image Diffusion Models Open
Text-to-image diffusion models have achieved remarkable success in generating high-quality contents from text prompts. However, their reliance on publicly available data and the growing trend of data sharing for fine-tuning make these mode…
View article: DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models
DiffExp: Efficient Exploration in Reward Fine-tuning for Text-to-Image Diffusion Models Open
Fine-tuning text-to-image diffusion models to maximize rewards has proven effective for enhancing model performance. However, reward fine-tuning methods often suffer from slow convergence due to online sample generation. Therefore, obtaini…
View article: Improbable Bigrams Expose Vulnerabilities of Incomplete Tokens in Byte-Level Tokenizers
Improbable Bigrams Expose Vulnerabilities of Incomplete Tokens in Byte-Level Tokenizers Open
Tokenization is a crucial step that bridges human-readable text with model-readable discrete tokens. However, recent studies have revealed that tokenizers can be exploited to elicit unwanted model behaviors. In this work, we investigate in…
View article: MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control
MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control Open
Autonomous agents powered by large language models (LLMs) show promising potential in assistive tasks across various domains, including mobile device control. As these agents interact directly with personal information and device settings,…
View article: Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation
Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation Open
Human feedback plays a critical role in learning and refining reward models for text-to-image generation, but the optimal form the feedback should take for learning an accurate reward function has not been conclusively established. This pa…
View article: Latent Action Pretraining from Videos
Latent Action Pretraining from Videos Open
We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised method for pretraining Vision-Language-Action (VLA) models without ground-truth robot action labels. Existing Vision-Language-Action models require ac…
View article: Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models
Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models Open
Fine-tuning text-to-image diffusion models with human feedback is an effective method for aligning model behavior with human intentions. However, this alignment process often suffers from slow convergence due to the large size and noise pr…
View article: DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing
DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing Open
Recent advances in diffusion models have introduced a new era of text-guided image manipulation, enabling users to create realistic edited images with simple textual prompts. However, there is significant concern about the potential misuse…
View article: Margin Matching Preference Optimization: Enhanced Model Alignment with Granular Feedback
Margin Matching Preference Optimization: Enhanced Model Alignment with Granular Feedback Open
Large language models (LLMs) fine-tuned with alignment techniques, such as reinforcement learning from human feedback, have been instrumental in developing some of the most capable AI systems to date. Despite their success, existing method…
View article: By My Eyes: Grounding Multimodal Large Language Models with Sensor Data via Visual Prompting
By My Eyes: Grounding Multimodal Large Language Models with Sensor Data via Visual Prompting Open
Large language models (LLMs) have demonstrated exceptional abilities across various domains. However, utilizing LLMs for ubiquitous sensing applications remains challenging as existing text-prompt methods show significant performance degra…
View article: Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation
Beyond Thumbs Up/Down: Untangling Challenges of Fine-Grained Feedback for Text-to-Image Generation Open
Human feedback plays a critical role in learning and refining reward models for text-to-image generation, but the optimal form the feedback should take for learning an accurate reward function has not been conclusively established. This pa…
View article: Aligning Large Language Models with Self-generated Preference Data
Aligning Large Language Models with Self-generated Preference Data Open
Aligning large language models (LLMs) with human preferences becomes a key component to obtaining state-of-the-art performance, but it yields a huge cost to construct a large human-annotated preference dataset. To tackle this problem, we p…
View article: Benchmarking Mobile Device Control Agents across Diverse Configurations
Benchmarking Mobile Device Control Agents across Diverse Configurations Open
Mobile device control agents can largely enhance user interactions and productivity by automating daily tasks. However, despite growing interest in developing practical agents, the absence of a commonly adopted benchmark in this area makes…
View article: Identity Decoupling for Multi-Subject Personalization of Text-to-Image Models
Identity Decoupling for Multi-Subject Personalization of Text-to-Image Models Open
Text-to-image diffusion models have shown remarkable success in generating personalized subjects based on a few reference images. However, current methods often fail when generating multiple subjects simultaneously, resulting in mixed iden…
View article: Confidence-aware Reward Optimization for Fine-tuning Text-to-Image Models
Confidence-aware Reward Optimization for Fine-tuning Text-to-Image Models Open
Fine-tuning text-to-image models with reward functions trained on human feedback data has proven effective for aligning model behavior with human intent. However, excessive optimization with such reward models, which serve as mere proxy ob…
View article: SelfReplay: Adapting Self-Supervised Sensory Models via Adaptive Meta-Task Replay
SelfReplay: Adapting Self-Supervised Sensory Models via Adaptive Meta-Task Replay Open
Self-supervised learning has emerged as a method for utilizing massive unlabeled data for pre-training models, providing an effective feature extractor for various mobile sensing applications. However, when deployed to end-users, these mod…