Article
Anthony Maio PRO
anthonym21
AI & ML interests
- Interpretability-oriented analysis of model behavior/welfare
- Heterogenous Model Swarms in Scalable Oversight / Automated Red Teaming
- Evaluation of internal consistency and failure modes
- Bio-mimetic memory and state management for long-running LLM agents
- Tooling to make agent behavior measurable and auditable
Organizations
Articles 6
Article
2
CoDA-GQA-L: Bounded-Memory Differential Attention with Value-Routed Landmark Banks
spaces 9
Running
Agents
Slipstream Lab
🧭
CPU-first technical companion for Slipstream 3.1.1
Running on Zero
Agents
CoDA-GQA-L Neural Database
🧠
Ingest documents, save neural states, and ask questions about them
Sleeping
Agents
Safety Lens
🐨
The Model MRI — a real-time activation scanner for HF models
Running
Agents
Ask My Research
🔬
Chat with Anthony Maio's AI safety research papers
Sleeping
Slipstream Governance Environment
🛡
Sleeping
Agents
World Model Demo
🧠
Watch an AI plan and move in a grid world
models 35
anthonym21/Eve-3-SABER-0.5B
0.5B • Updated
anthonym21/dots.ocr-GGUF
Image-to-Text • 2B • Updated • 103 • 2
anthonym21/eve-3-0.5b-saber
0.5B • Updated • 1
anthonym21/json-tokenizer-structured
Text Generation • Updated
anthonym21/Mistral-7B-v0.3-CoDA-GQA-L
Text Generation • 7B • Updated • 647
anthonym21/Eve-2-MoE-IT-272M
Text Generation • 0.3B • Updated • 1.22k • 1
anthonym21/Eve-2-MoE-272M
Text Generation • Updated • 589 • 1
anthonym21/Eve-2-MoE-NanoCommit-272M-GGUF
0.3B • Updated • 14
anthonym21/Eve-2-MoE-NanoPrompt-272M-GGUF
0.3B • Updated • 11
anthonym21/Eve-2-MoE-NanoSQL-272M-GGUF
0.3B • Updated • 14