Papers by Hirokazu Kiyomaru
Is a Knowledge-based Response Engaging?: An Analysis on Knowledge-Grounded Dialogue with Information Source Annotation (2023.acl-srw)
Copied to clipboard
| Challenge: | Currently, most knowledge-grounded dialogue models focus on reflecting given external knowledge. |
| Approach: | They analyze human behavior by annotating utterances in an existing knowledge-grounded dialogue corpus and find that speaker-derived information improves dialogue engagingness. |
| Outcome: | The proposed model cannot include speaker-derived information as often as humans do. |
Demystifying Mixed Outcomes of Self-Training: Pre-training Analyses on Non-Toy LLMs (2026.findings-eacl)
Copied to clipboard
| Challenge: | Recent studies on self-training report seemingly contradictory outcomes. |
| Approach: | They use OLMo-2 models as non-toy LLMs and perform multiple rounds of continual pre-training using self-generated text with different prompting strategies and data filtering. |
| Outcome: | The proposed model collapse is inherent to the training procedure itself, while self-improvement is likely owes its success to human-designed, strategic synthetic pipelines that inject external intelligence. |
Scaling Data-Constrained Language Models with Synthetic Data (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) improve with more training data, but practical limitations on data collection constrain further scaling. |
| Approach: | They compare three strategies to generate Japanese text, repeat the limited Japanese Web text, and use English Web text to fill the data shortfall. |
| Outcome: | The proposed model outperforms baselines and achieves the performance achieved when the entire token budget is filled with additional organic Japanese Web text. |
Diversity-aware Event Prediction based on a Conditional Variational Autoencoder with Reconstruction (D19-60)
Copied to clipboard
| Challenge: | Typical event sequences are important class of commonsense knowledge . previous work in event prediction uses sequence-to-sequence models . however, what can happen after a given event is usually diverse . |
| Approach: | They propose to incorporate a conditional variational autoencoder into seq2seq for its ability to represent diverse next events as a probabilistic distribution. |
| Outcome: | The proposed model outperforms deterministic models in terms of precision and recall . the proposed model is based on a conditional variational autoencoder . |
Abstractive Multi-Video Captioning: Benchmark Dataset Construction and Extensive Evaluation (2024.lrec-main)
Copied to clipboard
| Challenge: | Abstractive multi-video captioning focuses on abstracting multiple videos with natural language. |
| Approach: | They propose a task that generates an abstract caption of shared video content . they propose end-to-end and cascade approaches to abstractive multi-video captioning . |
| Outcome: | The proposed task generates an abstract caption of shared content in a video group containing multiple videos. |
Cross-lingual Knowledge Projection Using Machine Translation and Target-side Knowledge Base Completion (C18-1)
Copied to clipboard
| Challenge: | Existing efforts to build commonsense knowledge bases are expensive and lack quantity and quality between languages. |
| Approach: | They propose to project English commonsense knowledge into Japanese and Chinese with high precision. |
| Outcome: | The proposed method achieves top-10 accuracy on the crowdsourced English–Japanese benchmark and 18,747 facts of accurate Japanese commonsense within a very short period. |
MultiTool-CoT: GPT-3 Can Use Multiple External Tools with Chain of Thought Prompting (2023.acl-short)
Copied to clipboard
| Challenge: | Recent studies have focused on using a single external tool to solve a problem with large language models and have not addressed different problems together. |
| Approach: | They propose a framework that leverages chain-of-thought prompting to incorporate multiple external tools into the reasoning process. |
| Outcome: | The proposed framework outperforms baselines and achieves state-of-the-art performance on a task that requires both numerical reasoning and domain-specific knowledge. |
KWJA: A Unified Japanese Analyzer Based on Foundation Models (2023.acl-demo)
Copied to clipboard
Nobuhiro Ueda, Kazumasa Omura, Takashi Kodama, Hirokazu Kiyomaru, Yugo Murawaki, Daisuke Kawahara, Sadao Kurohashi
| Challenge: | KWJA supports a wide range of tasks including typo correction, word segmentation, word normalization, named entity recognition, dependency parsing, PAS analysis, bridging reference resolution, coreference resolution, and discourse relation analysis. |
| Approach: | They propose to build a Japanese text analyzer based on foundation models that performs a wide range of tasks. |
| Outcome: | The proposed model performs better in a multi-task manner than other analyzers with specialized models. |
Contextualized and Generalized Sentence Representations by Contrastive Self-Supervised Learning: A Case Study on Discourse Relation Analysis (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing methods to learn contextualized and generalized sentence representations are limited by the size of manually annotated data. |
| Approach: | They propose a method to learn contextualized and generalized sentence representations using contrastive self-supervised learning. |
| Outcome: | The proposed method outperforms baseline methods based on BERT, XLNet, and RoBERTa in English and Japanese and outperformed strong baseline methods. |
Constructing a Japanese Verdict Prediction Dataset for Fact-Checking of LLM-Generated Texts (2026.acl-srw)
Copied to clipboard
Miwa Masano, Hirokazu Kiyomaru, Atsushi Keyaki, Kaito Horio, Rei Minamoto, Ribeka Keyaki, Kouta Nakayama, Hideyuki Tachibana, Daisuke Kawahara
| Challenge: | Text generated by Large Language Models (LLMs) may contain plausible but incorrect information known as hallucinations. |
| Approach: | They extend the label set for verdict prediction to capture claim-evidence relationships humans would commonly interpret as supported or refuted. |
| Outcome: | The proposed system improves F1 by 4 percentage points compared to baseline. |