19 citations · 55 across the 25 of their papers we have counts for
28 papers
Do VLAs Really Follow Instructions? Diagnosing How-Level Instruction Following in Vision-Language-Action Models
Sangwu Park, Wonjoong Kim, Yeonjun In +1
Vision-language-action (VLA) models are expected to follow instructions like their language backbones, yet prior work has focused on What-level instructions that specify the task r…
ManiPhysicsBench: Physics-Based Assessment of Object Preservation in VLA Manipulation
Sangwu Park, Yeonjun In, Wonjoong Kim +3
Vision-language-action (VLA) models aim to perform diverse manipulation tasks, but task success in existing rigid-body benchmarks does not indicate whether they preserve objects. W…
Test-Time Training for Visual Foresight Vision-Language-Action Models
Sangwu Park, Wonjoong Kim, Yeonjun In +3
Visual Foresight VLA (VF-VLA) has become a prominent architectural choice in the recent VLA due to its impressive performance. Nevertheless, the inherent design of VF-VLA makes it…
Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents
Yeonjun In, Wonjoong Kim, Sangwu Park +2
Existing large language model (LLM) based memory systems apply universal, static policies that overlook a fundamental reality: the contexts that are worth storing in memory are dif…
PAIR: Prefix-Aware Internal Reward Model for Multi-Turn Agent Optimization
Wonjoong Kim, Yeonjun In, Sangwu Park +2
A significant hurdle for current LLMs is the execution of complex, multi-stage tasks. Group Relative Policy Optimization (GRPO) has been emerging as a leading choice, but its relia…
Reasoning Structure Matters for Safety Alignment of Reasoning Models
Yeonjun In, Wonjoong Kim, Sangwu Park +1
Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks but often generate harmful responses to malicious user queries. This paper investigates the unde…
Rethinking Failure Attribution in Multi-Agent Systems: A Multi-Perspective Benchmark and Evaluation
Yeonjun In, Mehrab Tanjim, Jayakumar Subramanian +6
Failure attribution is essential for diagnosing and improving multi-agent systems (MAS), yet existing benchmarks and methods largely assume a single deterministic root cause for ea…
Self-EvolveRec: Self-Evolving Recommender Systems with LLM-based Directional Feedback
Sein Kim, Sangwu Park, Hongseok Kang +6
Traditional methods for automating recommender system design, such as Neural Architecture Search (NAS), are often constrained by a fixed search space defined by human priors, limit…
Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents
Wonjoong Kim, Sangwu Park, Yeonjun In +3
Although recent tool-augmented benchmarks involve complex requests, evaluation remains limited to answer matching, neglecting critical trajectory aspects like efficiency, hallucina…
SelfJudge: Faster Speculative Decoding via Self-Supervised Judge Verification
Kanghoon Yoon, Minsub Kim, Sungjae Lee +6
Speculative decoding accelerates LLM inference by verifying candidate tokens from a draft model against a larger target model. Recent judge decoding boosts this process by relaxing…
R1-ACT: Efficient Reasoning Model Safety Alignment by Activating Safety Knowledge
Yeonjun In, Wonjoong Kim, Sangwu Park +1
Although large reasoning models (LRMs) have demonstrated impressive capabilities on complex tasks, recent studies reveal that these models frequently fulfill harmful user instructi…
Disambiguation in Conversational Question Answering in the Era of LLMs and Agents: A Survey
Md Mehrab Tanjim, Yeonjun In, Xiang Chen +8
Ambiguity remains a fundamental challenge in Natural Language Processing (NLP) due to the inherent complexity and flexibility of human language. With the advent of Large Language M…
Token-Efficient Item Representation via Images for LLM Recommender Systems
Kibum Kim, Sein Kim, Hongseok Kang +7
Large Language Models (LLMs) have recently emerged as a powerful backbone for recommender systems. Existing LLM-based recommender systems take two different approaches for represen…
Weakly Supervised Video Scene Graph Generation via Natural Language Supervision
Kibum Kim, Kanghoon Yoon, Yeonjun In +4
Existing Video Scene Graph Generation (VidSGG) studies are trained in a fully supervised manner, which requires all frames in a video to be annotated, thereby incurring high annota…
Training Robust Graph Neural Networks by Modeling Noise Dependencies
Yeonjun In, Kanghoon Yoon, Sukwon Yun +3
In real-world applications, node features in graphs often contain noise from various sources, leading to significant performance degradation in GNNs. Although several methods have…
Is Safety Standard Same for Everyone? User-Specific Safety Evaluation of Large Language Models
Yeonjun In, Wonjoong Kim, Kanghoon Yoon +5
As the use of large language model (LLM) agents continues to grow, their safety vulnerabilities have become increasingly evident. Extensive benchmarks evaluate various aspects of L…
RA-SGG: Retrieval-Augmented Scene Graph Generation Framework via Multi-Prototype Learning
Kanghoon Yoon, Kibum Kim, Jaehyung Jeon +3
Scene Graph Generation (SGG) research has suffered from two fundamental challenges: the long-tailed predicate distribution and semantic ambiguity between predicates. These challeng…
Revisiting Fake News Detection: Towards Temporality-aware Evaluation by Leveraging Engagement Earliness
Junghoon Kim, Junmo Lee, Yeonjun In +2
Social graph-based fake news detection aims to identify news articles containing false information by utilizing social contexts, e.g., user information, tweets and comments. Howeve…
Diversify-verify-adapt: Efficient and Robust Retrieval-Augmented Ambiguous Question Answering
Yeonjun In, Sungchul Kim, Ryan A. Rossi +4
The retrieval augmented generation (RAG) framework addresses an ambiguity in user queries in QA systems by retrieving passages that cover all plausible interpretations and generati…
Debiased Graph Poisoning Attack via Contrastive Surrogate Objective
Kanghoon Yoon, Yeonjun In, Namkyeong Lee +2
Graph neural networks (GNN) are vulnerable to adversarial attacks, which aim to degrade the performance of GNNs through imperceptible changes on the graph. However, we find that in…
SIMPLOT: Enhancing Chart Question Answering by Distilling Essentials
Wonjoong Kim, Sangwu Park, Yeonjun In +2
Recently, interpreting complex charts with logical reasoning has emerged as challenges due to the development of vision-language models. A prior state-of-the-art (SOTA) model has p…
DSLR: Diversity Enhancement and Structure Learning for Rehearsal-based Graph Continual Learning
Seungyoon Choi, Wonjoong Kim, Sungwon Kim +3
We investigate the replay buffer in rehearsal-based approaches for graph continual learning (GCL) methods. Existing rehearsal-based GCL methods select the most representative nodes…
Self-Guided Robust Graph Structure Refinement
Yeonjun In, Kanghoon Yoon, Kibum Kim +2
Recent studies have revealed that GNNs are vulnerable to adversarial attacks. To defend against such attacks, robust graph structure refinement (GSR) methods aim at minimizing the…
Adaptive Self-training Framework for Fine-grained Scene Graph Generation
Kibum Kim, Kanghoon Yoon, Yeonjun In +3
Scene graph generation (SGG) models have suffered from inherent problems regarding the benchmark datasets such as the long-tailed predicate distribution and missing annotation prob…