7 papers
Predicting LLM Reasoning Performance with Small Proxy Model
Woosung Koh, Juyoung Suk, Sungjun Han +2
Given the prohibitive cost of pre-training large language models, it is essential to leverage smaller proxy models to optimize datasets before scaling up. However, this approach be…
Generative Visual Code Mobile World Models
Woosung Koh, Sungjun Han, Segyu Lee +2
Mobile Graphical User Interface (GUI) World Models (WMs) offer a promising path for improving mobile GUI agent performance at train- and inference-time. However, current approaches…
Multi-Drafter Speculative Decoding with Alignment Feedback
Taehyeon Kim, Hojung Jung, Se-Young Yun
Speculative decoding (SD) accelerates large language model (LLM) inference by using a smaller model to draft future tokens, which are then verified by the target LLM. This preserve…
mSFT: Addressing Dataset Mixtures Overfitting Heterogeneously in Multi-task SFT
Woosung Koh, Jeyoung Jeon, Youngjin Song +4
Current language model training commonly applies multi-task Supervised Fine-Tuning (SFT) using a homogeneous compute budget across all sub-datasets. This approach is fundamentally…
AdaSTaR: Adaptive Data Sampling for Training Self-Taught Reasoners
Woosung Koh, Wonbeen Oh, Jaein Jang +7
Self-Taught Reasoners (STaR), synonymously known as Rejection sampling Fine-Tuning (RFT), is an integral part of the training pipeline of self-improving reasoning Language Models (…
FlickerFusion: Intra-trajectory Domain Generalizing Multi-Agent RL
Woosung Koh, Wonbeen Oh, Siyeol Kim +5
Multi-agent reinforcement learning has demonstrated significant potential in addressing complex cooperative tasks across various real-world applications. However, existing MARL app…