_rootcomputer

Rootcomputer is an independent AI research lab focused on understanding, building, and evaluating modern machine learning systems.

Our work centers on small and mid-scale language models, efficient transformer architectures, training dynamics, alignment, model behavior, and the internal representations that emerge during training.

We are especially interested in how architectural choices, data design, and optimization strategy affect reliability, reasoning, grounding, and failure modes in neural networks.

Research Focus

Rootcomputer conducts applied and exploratory research across several areas:

• Efficient transformer-based model development
• Small and mid-scale language model training
• Alignment and behavioral stability through architecture and data
• Training dynamics, curriculum design, and optimization behavior
• Failure analysis, robustness, and generalization limits
• Curated corpora for grounding, reasoning, and conversational stability
• Specialist and narrow-domain AI systems
• Internal representation, cognition, and emergent behavior in neural networks

Small Models First

Rootcomputer is built around the belief that meaningful AI research does not require scale alone.

Small and mid-scale models provide a controllable and experimentally rich environment for studying how models learn, fail, generalize, and align. By constraining size, we hypothesize that we will gain some clarity into the mechanics of training and behavior that are often harder to isolate in very large systems.

Models

Haiku Mini

Haiku Mini is a compact conversational research model focused on grounding, and behavioral consistency in small language models.

It is designed as a controlled platform for studying turn-taking, instruction following, hallucination, and conversational stability under tight parameter constraints.

Tanka

Tanka extends the Haiku research line into simple reasoning, more consistant factual recall, utilizing about twice as many total parameters as Haiku Mini.

It is intended for internal research, evaluation, and architectural experimentation. Limited public access due to the nature of an experimental project.

Data & Training

Rootcomputer treats training data as a first-class research object.

Our datasets and corpora are designed to study:

• Conversational grounding
• Response finality
• Instruction following
• Reasoning behavior
• Ethical constraint learning
• Specialist-domain behavior
• Alignment under limited scale

We also build custom training software and infrastructure for controlled experimentation, including transformer training pipelines, reproducible evaluation, and limited-hardware friendly solutions.

Safety & Alignment

Safety and alignment are central to Rootcomputer’s research direction.

We study alignment as a property shaped by architecture, data, training dynamics, evaluation pressure, and deployment constraints. Our work emphasizes predictable behavior, controlled interaction, failure-mode discovery, and transparent model limitations.

Links

• Website: https://rootcomputer.dev
• Haiku Mini: https://chathaiku.com
• Hugging Face: https://huggingface.co/rootcomputer

Organization

Rootcomputer AI Development Branch