top of page

Search

The AI Meeting Nobody Needed

The AI Meeting Nobody Needed

The AI Meeting Nobody Needed

5 min read

The 14% revenue opportunity hiding inside one rate card

The 14% revenue opportunity hiding inside one rate card

The 14% revenue opportunity hiding inside one rate card

6 min read

Jev Beat the 4B Model. Then the 4B Model Beat Jev.

Jev Beat the 4B Model. Then the 4B Model Beat Jev.

Jev Beat the 4B Model. Then the 4B Model Beat Jev.

11 min read

Why Couldn't an LLM Pick Its Best Expert?

Why Couldn't an LLM Pick Its Best Expert?

Why Couldn't an LLM Pick Its Best Expert?

6 min read

Why thin-file approvals look fine in backtest and fail in production

Why thin-file approvals look fine in backtest and fail in production

Why thin-file approvals look fine in backtest and fail in production

4 min read

Your fraud model has never seen the attack that will work

Your fraud model has never seen the attack that will work

Your fraud model has never seen the attack that will work

4 min read

Why construction estimates lose their margin after the bid is won

Why construction estimates lose their margin after the bid is won

Why construction estimates lose their margin after the bid is won

4 min read

What travels around the loop?

What travels around the loop?

What travels around the loop?

7 min read

Deleting a Model's Thoughts Changed Nothing. Then I Changed the Task.

Deleting a Model's Thoughts Changed Nothing. Then I Changed the Task.

Deleting a Model's Thoughts Changed Nothing. Then I Changed the Task.

9 min read

I Sent a Small AI Model to School. Here’s Its Report Card.

I Sent a Small AI Model to School. Here’s Its Report Card.

I Sent a Small AI Model to School. Here’s Its Report Card.

5 min read

I forced a Mixture of Experts to specialize by topic. It got worse.

I forced a Mixture of Experts to specialize by topic. It got worse.

I forced a Mixture of Experts to specialize by topic. It got worse.

9 min read

What does a Mixture-of-Experts router actually read?

What does a Mixture-of-Experts router actually read?

What does a Mixture-of-Experts router actually read?

5 min read

Mixture of Experts: the sparse model that lights up 93% of itself

Mixture of Experts: the sparse model that lights up 93% of itself

Mixture of Experts: the sparse model that lights up 93% of itself

6 min read

J-Scope: before a model answers, it thinks in concepts

J-Scope: before a model answers, it thinks in concepts

J-Scope: before a model answers, it thinks in concepts

7 min read

I Fed a Model Facts Through Its J-Space Instead of Its Prompt

I Fed a Model Facts Through Its J-Space Instead of Its Prompt

I Fed a Model Facts Through Its J-Space Instead of Its Prompt

12 min read

Can you tell a good AI agent plan from a bad one before you run it? I spent 6 million tokens finding out.

Can you tell a good AI agent plan from a bad one before you run it? I spent 6 million tokens finding out.

Can you tell a good AI agent plan from a bad one before you run it? I spent 6 million tokens finding out.

8 min read

I Gave My Coding Agent Karpathy's Discipline Rules. It Got Too Careful to Fix the Bug.

I Gave My Coding Agent Karpathy's Discipline Rules. It Got Too Careful to Fix the Bug.

I Gave My Coding Agent Karpathy's Discipline Rules. It Got Too Careful to Fix the Bug.

6 min read

I bolted MiniMax's MSA sparse attention onto a 3B model on a single 4090

I bolted MiniMax's MSA sparse attention onto a 3B model on a single 4090

I bolted MiniMax's MSA sparse attention onto a 3B model on a single 4090

10 min read

I ran DFlash on a MacBook. The 4x is a data-center number.

I ran DFlash on a MacBook. The 4x is a data-center number.

I ran DFlash on a MacBook. The 4x is a data-center number.

7 min read

Why your adaptive controller might be causing the defects it’s trying to prevent

Why your adaptive controller might be causing the defects it’s trying to prevent

Why your adaptive controller might be causing the defects it’s trying to prevent

3 min read

Recent Insights

bottom of page