4 papers
Relevance to Utility: Process-Supervised Rewrite for RAG
Jaeyoung Kim, Jongho Kim, Seung-won Hwang +2
Retrieval-augmented generation systems often suffer from a gap between optimizing retrieval relevance and generative utility. With such a gap, retrieved documents may be topically…
Intended Target Identification for Anomia Patients with Gradient-based Selective Augmentation
Jongho Kim, Romain Storaï, Seung-won Hwang
In this study, we investigate the potential of language models (LMs) in aiding patients experiencing anomia, a difficulty identifying the names of items. Identifying the intended t…
Counterfactual-Consistency Prompting for Relative Temporal Understanding in Large Language Models
Jongho Kim, Seung-won Hwang
Despite the advanced capabilities of large language models (LLMs), their temporal reasoning ability remains underdeveloped. Prior works have highlighted this limitation, particular…
HARP: Hesitation-Aware Reframing in Transformer Inference Pass
Romain Storaï, Seung-won Hwang
This paper aims to improve the performance of large language models by addressing the variable computational demands in inference steps, where some tokens require more computationa…