activity
20242026
collaborators

14 papers

cs.CL2026

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…

cs.CL2026

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…

cs.CL2026

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…

cs.CL2026

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

Yusuke Takase, Momose Oyama, 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…

cs.CL2026

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…

cs.CL2026

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…