arXiv subjects

Yusuke Yamauchi

Publications and source records attributed to Yusuke Yamauchi.

5 recordsLinked to original sources

arXiv2026

Are Emotions Arranged in a Circle? Geometric Analysis of Emotion Representations via Hyperspherical Contrastive Learning

Psychological research has long utilized circumplex models to structure emotions, placing similar emotions adjacently and opposing ones diagonally. Although frequently used to interpret deep learning representations, these models are rarely directly incorporated into the representation learning of language models, leaving their geometric validity unexplored. This paper proposes a method to induce circular emotion representations within language model embeddings via contrastive learning on a hypersphere. We show that while this circular alignment offers superior interpretability and robustness against dimensionality reduction, it underperforms compared to conventional designs in high-dimensional settings and fine-grained classification. Our findings elucidate the trade-offs involved in applying psychological circumplex models to deep learning architectures.

cs.CL↗

arXiv2025

Massive Supervised Fine-tuning Experiments Reveal How Data, Layer, and Training Factors Shape LLM Alignment Quality

Supervised fine-tuning (SFT) is a critical step in aligning large language models (LLMs) with human instructions and values, yet many aspects of SFT remain poorly understood. We trained a wide range of base models on a variety of datasets including code generation, mathematical reasoning, and general-domain tasks, resulting in 1,000+ SFT models under controlled conditions. We then identified the dataset properties that matter most and examined the layer-wise modifications introduced by SFT. Our findings reveal that some training-task synergies persist across all models while others vary substantially, emphasizing the importance of model-specific strategies. Moreover, we demonstrate that perplexity consistently predicts SFT effectiveness, often surpassing superficial similarity between the training data and the benchmark, and that mid-layer weight changes correlate most strongly with performance gains. We release these 1,000+ SFT models and benchmark results to accelerate further research. All resources are available at https://github.com/llm-jp/massive-sft.

cs.CL↗

arXiv2025

An Empirical Study of LLM-as-a-Judge: How Design Choices Impact Evaluation Reliability

As large language models (LLMs) continue to advance, reliable evaluation methods are essential particularly for open-ended, instruction-following tasks. LLM-as-a-Judge enables automatic evaluation using LLMs as evaluators, but its reliability remains uncertain. In this work, we analyze key factors affecting its trustworthiness, focusing on alignment with human judgments and evaluation consistency. Using BIGGENBench and EvalBiasBench, we study the effects of evaluation design, decoding strategies, and Chain-of-Tought (CoT) reasoning in evaluation. Our results show that evaluation criteria are critical for reliability, non-deterministic sampling improves alignment with human preferences over deterministic evaluation, and CoT reasoning offers minimal gains when clear evaluation criteria are present.

cs.CL↗

arXiv2024

LLM-jp: A Cross-organizational Project for the Research and Development of Fully Open Japanese LLMs

This paper introduces LLM-jp, a cross-organizational project for the research and development of Japanese large language models (LLMs). LLM-jp aims to develop open-source and strong Japanese LLMs, and as of this writing, more than 1,500 participants from academia and industry are working together for this purpose. This paper presents the background of the establishment of LLM-jp, summaries of its activities, and technical reports on the LLMs developed by LLM-jp. For the latest activities, visit https://llm-jp.nii.ac.jp/en/.

cs.CL↗

arXiv2022

Plasmon triggered ultrafast operation of color centers in hBN layers

High-quality emission centers in two-dimensional materials are promising components for future photonic and optoelectronic applications. Carbon-enriched hexagonal boron nitride (hBN:C) layers host atom-like color-center (CC) defects with strong and robust photoemission up to room temperature. Placing the hBN:C layers on top of Ag triangle nanoparticles (NPs) accelerate the decay of the CC defects down to 46 ps from their reference bulk value of 350 ps. The ultrafast decay is achieved due to the efficient excitation of the plasmon modes of the Ag NPs by the near field of the CCs. Simulations of the CCs/Ag NP interaction present that higher Purcell values are expected, although the measured decay of the CCs is limited by the instrument response. The influence of the NP thickness to the Purcell factor of the CCs is analyzed. The ultrafast operation of the CCs in hBN:C layers paves the way for their use in demanding applications, such as single-photon emitters and quantum devices.

physics.optics↗