H. Shimodaira

My FAULT: Self-Diagnosis as Credit Assignment in Self-Evolving Agentic Reinforcement Learning

Yihua Zhu, Qianying Liu, Weixu Qiao +10

Agentic reinforcement learning (RL) has emerged as a powerful approach for training large language model agents on multi-step tasks, yet reliance on terminal outcome rewards create…

Reasoning Depth and Environment Complexity: A Controlled Study of RLVR Data Allocation across Logical Reasoning Tasks

Yihua Zhu, Qianying Liu, Fei Cheng +4

Reinforcement learning with verifiable rewards (RLVR) has become central to post-training reasoning models, yet a key limitation of existing studies is their narrow view of the rea…

Language Model Maps for Prompt-Response Distributions via Log-Likelihood Vectors

Momose Oyama, Yusuke Takase, Hidetoshi Shimodaira

We propose a method that represents language models by log-likelihood vectors over prompt-response pairs and constructs model maps for comparing their conditional distributions. In…

Domain Mixture Design via Log-Likelihood Differences for Aligning Language Models with a Target Model

Ryo Kishino, Riku Shiomi, Hiroaki Yamagiwa +2

Instead of directly distilling a language model, this study addresses the problem of aligning a base model with a target model in distribution by designing the domain mixture of tr…

Measuring Affinity between Attention-Head Weight Subspaces via the Projection Kernel

Hiroaki Yamagiwa, Yusuke Takase, Hidetoshi Shimodaira

Understanding relationships between attention heads is essential for interpreting the internal structure of Transformers, yet existing metrics do not capture this structure well. W…

Memorization, Emergence, and Explaining Reversal Failures: A Controlled Study of Relational Semantics in LLMs

Yihua Zhu, Qianying Liu, Jiaxin Wang +5

Autoregressive LLMs perform well on relational tasks that require linking entities via relational words (e.g., father/son, friend), but it is unclear whether they learn the logical…

Likelihood Variance as Text Importance for Resampling Texts to Map Language Models

Momose Oyama, Ryo Kishino, Hiroaki Yamagiwa +1

We address the computational cost of constructing a model map, which embeds diverse language models into a common space for comparison via KL divergence. The map relies on log-like…

Establishing a Scale for Kullback-Leibler Divergence in Language Models Across Various Settings

Ryo Kishino, Yusuke Takase, Momose Oyama +2

Log-likelihood vectors define a common space for comparing language models as probability distributions, enabling unified comparisons across heterogeneous settings. We extend this…

Beyond Chains: Bridging Large Language Models and Knowledge Bases in Complex Question Answering

Yihua Zhu, Qianying Liu, Akiko Aizawa +1

Knowledge Base Question Answering (KBQA) aims to answer natural language questions using structured knowledge from KBs. While LLM-only approaches offer generalization, they suffer…

DeLTa: A Decoding Strategy based on Logit Trajectory Prediction Improves Factuality and Reasoning Ability

Yunzhen He, Yusuke Takase, Yoichi Ishibashi +1

Large Language Models (LLMs) are increasingly being used in real-world applications. However, concerns about the reliability of the content they generate persist, as it frequently…

Mapping 1,000+ Language Models via the Log-Likelihood Vector

Momose Oyama, Hiroaki Yamagiwa, Yusuke Takase +1

To compare autoregressive language models at scale, we propose using log-likelihood vectors computed on a predefined text set as model features. This approach has a solid theoretic…

Quantifying Lexical Semantic Shift via Unbalanced Optimal Transport

Ryo Kishino, Hiroaki Yamagiwa, Ryo Nagata +2

Lexical semantic change detection aims to identify shifts in word meanings over time. While existing methods using embeddings from a diachronic corpus pair estimate the degree of c…

Understanding Higher-Order Correlations Among Semantic Components in Embeddings

Momose Oyama, Hiroaki Yamagiwa, Hidetoshi Shimodaira

Independent Component Analysis (ICA) offers interpretable semantic components of embeddings. While ICA theory assumes that embeddings can be linearly decomposed into independent co…

Norm of Mean Contextualized Embeddings Determines their Variance

Hiroaki Yamagiwa, Hidetoshi Shimodaira

Contextualized embeddings vary by context, even for the same token, and form a distribution in the embedding space. To analyze this distribution, we focus on the norm of the mean e…

Shimo Lab at "Discharge Me!": Discharge Summarization by Prompt-Driven Concatenation of Electronic Health Record Sections

Yunzhen He, Hiroaki Yamagiwa, Hidetoshi Shimodaira

In this paper, we present our approach to the shared task "Discharge Me!" at the BioNLP Workshop 2024. The primary goal of this task is to reduce the time and effort clinicians spe…