AI Intelligence Archive

Image enhancement and restoration have become standard back-end operations on short-video and social media platforms to boost UGC visual experience. Yet these processes inevitably introduce visual anomalies--especially in faces, texts, and…

Every serious engineering organization has spent decades building governance around one artifact:...

Large Language Models (LLMs) are increasingly used to predict the responses of human participants in survey panels. Towards that goal, persona prompting has recently emerged as a technique to inform and align large pretrained language mode…

Have you ever read a job posting and your heart just sings the song of "HIRE ME PLEASE!!!", yeah,...

Uncertain knowledge graphs (UKGs) extend knowledge graphs by assigning each triple a continuous confidence score. Since most possible triples lack observed confidences, recent methods rely on semi-supervised learning to generate pseudo-lab…

Before the world knew it was a product of Z.ai, a massive Chinese tech player, this model was running...

here's a meme we've all posted some version of. "My code works and I have no idea why." Classic....

We study body-only, 12-class acted-emotion classification from skeleton motion under leave-performer-out (LPO) evaluation, a hard, underdetermined setting: chance is 8.3%, and a protocol-matched reproduced STGCN++ baseline reaches only 25.…

Test-Time Adaptation (TTA) has recently emerged as a promising strategy that allows the adaptation of pre-trained models to changing data distributions at deployment time, without access to any labels. To mitigate error accumulation, resea…

For the last month, every AI coding session on this machine ran under grith, which intercepts every...

Text-to-image (T2I) models have achieved remarkable success at faithfully rendering specified objects and attributes, yet their ability to produce visual metaphors, images that convey abstract ideas by combining elements from two distinct …

Recommender-system experiments often rely on a single random training seed, assuming that run-to-run stochasticity has limited impact on evaluation conclusions. This assumption is risky, as a training seed may influence several algorithm-d…

A real estate website can generate a buyer inquiry in seconds, but one wrong chatbot answer can undo...

Zero-shot graph anomaly detection seeks to deploy a detector trained on source graphs to unseen, unlabeled targets, yet domain shift can make source-derived notions of normality unreliable. We introduce RINSE (Robust Iterative Normality Se…

My last article shared the high-level architecture of KODA (my AI coding mentor built on a POCO C55)....

If you are using AI coding assistants like Cursor, Windsurf, Claude Desktop, or ChatGPT, you have...

TL;DR I learned and applied Open Knowledge Format (OKF), a vendor-neutral Markdown-and-YAML format...

Why tool scoping matters more than anything the model is told Part 7 findings of an experiment:...

Something strange has been happening lately, I keep forgetting small things. Not names. Not...

While making social media creatives for clients. AI tools design them fine now, but the output is one...

You are likely using Cursor wrong. Here is how to set boundaries for the AI so it stops hallucinating your architecture and starts delivering usable code.

The junior's pull request appeared fantastic. They had made clean commits, maintained a tidy...

A worked example of the frontmatter contract, and a short threat model for agents that are allowed to call tools.

Previously: 9 Bugs That All Looked Like a Working System · I Built an AI That Rewrites Its Own...

Three real patterns for gating an agent before it does something irreversible, drawn from projects I built.

Here's a fact that surprises people new to Kubernetes and should genuinely alarm them a little: by...

Previously: 9 Bugs That All Looked Like a Working System · I Built an AI That Rewrites Its Own...

There is a ceiling on your test coverage and it is not technical. Running tests has been a solved...

My publish-pass agent hit a beat that no source label could satisfy — every relabel that cleared one automated gate tripped another, provably, by arithmetic. The fix wasn't to relabel harder; it was a recorded, attributable waiver, reserve…

I run Claude Code across eleven projects at the same time. Not eleven files. Eleven separate...

Every week someone asks which coding agent is better, and every answer is a vibe. I wanted a...

I have been building an AI-native PCB editor for a few months. Before I wrote a line of it, I did the...

Undo almost never proves what it undid Falsifier, up front: if you can point me to a real,...

Every additional agent in a chain is supposed to make the system smarter. Too often, it just makes...

Agents are getting wired to money — order food, top up credits, call paid APIs. Between 'the agent wants to spend' and 'the money moves' you need a decision layer: ALLOW / human approval / DENY, with audit.

A reported Claude Fable 5.1 prompt extraction reveals why production AI agents depend on tools, retrieval, memory policies, provenance, and application-level safeguards.

Six weeks after I fixed capture with hooks, my second brain was still confidently wrong. The nightly...

On September 1 my upstream provider repriced one of the models I depend on. Here's the part of the...

When I started in this field, I quietly believed there would come a point where I had learned enough....