Building on HF

Brian Moran PRO

obversarystudios

obversarystu
architectfromthefuture
brian-moran-89a0a4406

AI & ML interests

AI evaluation, agent observability, memory substrates, failure traces, boundary mapping, prompt-injection defense, constrained policy learning, cognitive systems

Recent Activity

published a Space 4 days ago

obversarystudios/deepseek-ai-DeepSeek-V4-Pro

liked a model 4 days ago

deepseek-ai/DeepSeek-V4-Pro

upvoted an article 4 days ago

DeepSeek-V4: a million-token context that agents can actually use

View all activity

Organizations

None yet

obversarystudios 's Spaces 5

Deepseek Ai DeepSeek V4 Pro

🌖

Chat with DeepSeek V4 Pro AI model

Agent Threat Map Observatory

🧭

Threat-map benchmark with metrics and geometry

Failure Geometry Demo

🧩

Failure structure discovery on CARB reasoning

CARB Failure Observability

🔬

Failure analysis for LM reasoning via HF Inference API

Bringing paper to life: A modern template for scientific writing

📝