8 citations · 16 across the 19 of their papers we have counts for
22 papers
LLM-as-an-Improver: Turning Verification into Better Candidates
Akiyoshi Tomihari, Yuma Ichikawa
Verifier-based selection improves LLM performance by generating multiple candidate solutions and using a verifier to select the most promising one. However, existing methods typica…
Kozuchi Agent: A Language-Agnostic Open-Weight Agent for Software Repair
Mehdi Bahrami, Kosaku Kimura, Satoshi Munakata +24
Industrial software-engineering teams increasingly need LLM agents that turn bug reports into correct patches, yet benchmark-scale operation adds long horizons, tool-use discipline…
One Qubit Can Beat One Bit: Quantum Advantage for Post-Training Quantization
Yuma Ichikawa, Moeto Mishima
One-bit post-training quantization represents each weight using only its sign, requiring all deployment contexts to share the same binary weight matrix even when their activation s…
Addressable Recall Compaction for Long Context-Window Control in AI Agents
Thang Dang, Yuma Ichikawa, Sakina Fatima +1
Long-horizon LLM agents accumulate reasoning traces, actions, and tool observations that can eventually exceed a model's fixed context window. Existing compaction methods address t…
LOGOS: A Living Logic for AI Agent Teams That Evolve With Humans
Yuma Ichikawa, Yamato Arai, Kosaku Kimura +2
AI agents are evolving from answer engines into persistent teams that use tools, delegate work, learn from experience, and modify the artifacts that shape their future behavior. Th…
EVE-Agent: Evidence-Verifiable Self-Evolving Agents
Yamato Arai, Yuma Ichikawa
Self-evolving agents should not train on examples they cannot justify. Data-free self-evolving search agents offer a scalable route to systems that generate their own questions, an…
Tensor-Network Population Annealing
Takumi Oshima, Yuma Ichikawa, Koji Hukushima
We propose a hybrid sampling method, tensor-network population annealing (TNPA), which combines tensor-network (TN) initialization with population annealing (PA). We apply this met…
OneComp: One-Line Revolution for Generative AI Model Compression
Yuma Ichikawa, Keiji Kimura, Akihiro Yoshida +11
Deploying foundation models is increasingly constrained by memory footprint, latency, and hardware costs. Post-training compression can mitigate these bottlenecks by reducing the p…
Thermal Min-Max Games: Unifying Bounded Rationality and Typical-Case Equilibrium
Yuma Ichikawa
Strategic-form min-max game theory examines the existence, multiplicity, selection of equilibria, and the worst-case computational complexity under perfect rationality. However, in…
High-Dimensional Learning Dynamics of Quantized Models with Straight-Through Estimator
Yuma Ichikawa, Shuhei Kashiwamura, Ayaka Sakata
Quantized neural network training optimizes a discrete, non-differentiable objective. The straight-through estimator (STE) enables backpropagation through surrogate gradients and i…
Quantization Error Propagation: Revisiting Layer-Wise Post-Training Quantization
Yamato Arai, Yuma Ichikawa
Layer-wise PTQ is a promising technique for compressing large language models (LLMs), due to its simplicity and effectiveness without requiring retraining. However, recent progress…
Optimization by Parallel Quasi-Quantum Annealing with Gradient-Based Sampling
Yuma Ichikawa, Yamato Arai
Learning-based methods have gained attention as general-purpose solvers due to their ability to automatically learn problem-specific heuristics, reducing the need for manually craf…
Ratio Divergence Learning Using Target Energy in Restricted Boltzmann Machines: Beyond Kullback--Leibler Divergence Learning
Yuichi Ishida, Yuma Ichikawa, Aki Dote +2
We propose ratio divergence (RD) learning for discrete energy-based models, a method that utilizes both training data and a tractable target energy function. We apply RD learning t…
Statistical Mechanics of Min-Max Problems
Yuma Ichikawa, Koji Hukushima
Min-max optimization problems, also known as saddle point problems, have attracted significant attention due to their applications in various fields, such as fair beamforming, gene…
Training-Free Time-Series Anomaly Detection: Leveraging Image Foundation Models
Nobuo Namura, Yuma Ichikawa
Recent advancements in time-series anomaly detection have relied on deep learning models to handle the diverse behaviors of time-series data. However, these models often suffer fro…
Continuous Parallel Relaxation for Finding Diverse Solutions in Combinatorial Optimization Problems
Yuma Ichikawa, Hiroaki Iwashita
Finding the optimal solution is often the primary goal in combinatorial optimization (CO). However, real-world applications frequently require diverse solutions rather than a singl…
Adaptive Flip Graph Algorithm for Matrix Multiplication
Yamato Arai, Yuma Ichikawa, Koji Hukushima
This study proposes the "adaptive flip graph algorithm", which combines adaptive searches with the flip graph algorithm for finding fast and efficient methods for matrix multiplica…
Learning Dynamics in Linear VAE: Posterior Collapse Threshold, Superfluous Latent Space Pitfalls, and Speedup with KL Annealing
Yuma Ichikawa, Koji Hukushima
Variational autoencoders (VAEs) face a notorious problem wherein the variational posterior often aligns closely with the prior, a phenomenon known as posterior collapse, which hind…
Controlling Continuous Relaxation for Combinatorial Optimization
Yuma Ichikawa
Unsupervised learning (UL)-based solvers for combinatorial optimization (CO) train a neural network that generates a soft solution by directly optimizing the CO objective using a c…
High-dimensional Asymptotics of VAEs: Threshold of Posterior Collapse and Dataset-Size Dependence of Rate-Distortion Curve
Yuma Ichikawa, Koji Hukushima
In variational autoencoders (VAEs), the variational posterior often collapses to the prior, known as posterior collapse, which leads to poor representation learning quality. An adj…
Toward Unlimited Self-Learning MCMC with Parallel Adaptive Annealing
Yuma Ichikawa, Akira Nakagawa, Hiromoto Masayuki +1
Self-learning Monte Carlo (SLMC) methods are recently proposed to accelerate Markov chain Monte Carlo (MCMC) methods using a machine learning model. With latent generative models,…
Statistical-mechanical study of deep Boltzmann machine given weight parameters after training by singular value decomposition
Yuma Ichikawa, Koji Hukushima
Deep learning methods relying on multi-layered networks have been actively studied in a wide range of fields in recent years, and deep Boltzmann machines(DBMs) is one of them. In t…