collaborators

6 papers

stat.ML2026

How Neural Reward Models Learn Features for Policy Optimization: A Single-Index Analysis

Rei Higuchi, Ryotaro Kawata, Akifumi Wachi +3

Reward modeling is not only a prediction problem: in KL-regularized policy optimization, the learned reward is exponentiated to define the deployed policy, so downstream value depe…

stat.ML2026

Transformers as Measure-Theoretic Associative Memory: A Statistical Perspective and Minimax Optimality

Ryotaro Kawata, Taiji Suzuki

Transformers excel through content-addressable retrieval and the ability to exploit contexts of, in principle, unbounded length. We recast associative memory at the level of probab…

cs.LG2025

From Shortcut to Induction Head: How Data Diversity Shapes Algorithm Selection in Transformers

Ryotaro Kawata, Yujin Song, Alberto Bietti +4

Transformers can implement both generalizable algorithms (e.g., induction heads) and simple positional shortcuts (e.g., memorizing fixed output positions). In this work, we study h…

cs.LG2025

Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning

Ryotaro Kawata, Kohsei Matsutani, Yuri Kinoshita +2

Mixture of Experts (MoE), an ensemble of specialized models equipped with a router that dynamically distributes each input to appropriate experts, has achieved successful results i…

cs.CL2025

When Does Metadata Conditioning (NOT) Work for Language Model Pre-Training? A Study with Context-Free Grammars

Rei Higuchi, Ryotaro Kawata, Naoki Nishikawa +7

The ability to acquire latent semantics is one of the key properties that determines the performance of language models. One convenient approach to invoke this ability is to prepen…

cs.LG2025

Direct Distributional Optimization for Provable Alignment of Diffusion Models

Ryotaro Kawata, Kazusato Oko, Atsushi Nitanda +1

We introduce a novel alignment method for diffusion models from distribution optimization perspectives while providing rigorous convergence guarantees. We first formulate the probl…