6 papers
Dual-Scale World Models for LLM Agents Towards Hard-Exploration Problems
Minsoo Kim, Seung-won Hwang
LLM-based agents have seen promising advances, yet they are still limited in "hard-exploration" tasks requiring learning new knowledge through exploration. We present GLoW, a novel…
Agent-as-Judge for Factual Summarization of Long Narratives
Yeonseok Jeong, Minsoo Kim, Seung-won Hwang +1
Large Language Models (LLMs) have demonstrated near-human performance in summarization tasks based on traditional metrics such as ROUGE and BERTScore. However, these metrics do not…
CoEx -- Co-evolving World-model and Exploration
Minsoo Kim, Seung-won Hwang
Planning in modern LLM agents relies on the utilization of LLM as an internal world model, acquired during pretraining. However, existing agent designs fail to effectively assimila…
Chaining Event Spans for Temporal Relation Grounding
Jongho Kim, Dohyeon Lee, Minsoo Kim +1
Accurately understanding temporal relations between events is a critical building block of diverse tasks, such as temporal reading comprehension (TRC) and relation extraction (TRE)…
ECoRAG: Evidentiality-guided Compression for Long Context RAG
Yeonseok Jeong, Jinsu Kim, Dohyeon Lee +1
Large Language Models (LLMs) have shown remarkable performance in Open-Domain Question Answering (ODQA) by leveraging external documents through Retrieval-Augmented Generation (RAG…
From Token to Action: State Machine Reasoning to Mitigate Overthinking in Information Retrieval
Dohyeon Lee, Yeonseok Jeong, Seung-won Hwang
Chain-of-Thought (CoT) prompting enables complex reasoning in large language models (LLMs), including applications in information retrieval (IR). However, it often leads to overthi…