Michael Mathias

Data doings!

[berrybyrivabit @berrybyrivabit]

The wait is over!!!!

“Berry by Rivabit” is live on the AppStore now. Download the Berry App today. Be a part of our first draw!

50,000 NGN up for this weekend

https://t.co/BWjFpYqFhb

#berryislive #berrybyrivabit #berryishere

[berrybyrivabit @berrybyrivabit]

The wait is over!!!!

“Berry by Rivabit” is live on the AppStore now. Download the Berry App today. Be a part of our first draw!

50,000 NGN up for this weekend

https://t.co/BWjFpYqFhb

#berryislive #berrybyrivabit #berryishere

[Instablog9ja @instablog9ja]

“I want to be remembered as someone who has industrialised Africa” — Billionaire Aliko Dangote

[berrybyrivabit @berrybyrivabit]

UPDATE ‼️‼️

Launch date has been moved!

#berryiscooking #berryiscoming

[berrybyrivabit @berrybyrivabit]

Meet Berries.

Earn them.
Save them.
Cash them out.
Or use them for a chance to win bigger rewards.

Simple.

3 days to go.

#berryiscoming #berrybyrivabit

[berrybyrivabit @berrybyrivabit]

Imagine hearing about the app that pays people for participating...
after everyone else has made the most of it.

Yeah.

Don’t let that be you.

2 days.

#berryiscoming #berrybyrivabit

[berrybyrivabit @berrybyrivabit]

What if every survey completed, every task finished,
every minute spent engaging...
actually earned you something?

4 days.

Something rewarding is coming.

#berryiscoming #berrybyrivabit

[berrybyrivabit @berrybyrivabit]

Get familiar with the currency, Berries

#berrybyrivabit #berryinbloom #berryiscoming

[berrybyrivabit @berrybyrivabit]

Genuine question.

You spend hours on your phone every day. Brands make billions from your attention.

How much of that gets in you pocket?

Something is coming. Follow Berrybyrivabit find out.

#berryinbloom #berrybyrivabit #earnberries

[berrybyrivabit @berrybyrivabit]

Your time is valuable.Your opinions are valuable.Your attention is valuable.

So why are you giving it away for free? Something is changing.

Follow @berrybyrivabit to find out when.

#berriesinbloom #berrybyrivabit #earnberries

[berrybyrivabit @berrybyrivabit]

Something new is coming to Nigeria. Something that actually puts money back in your pocket. Follow. You’ll want to see this.

#berriesinbloom #berrybyrivabit #earnberries

[berrybyrivabit @berrybyrivabit]

Berry is coming

#berriesinbloom #berrybyrivabit

[berrybyrivabit @berrybyrivabit]

Studies show that people pick up their phones about 50 times daily on an average…

[berrybyrivabit @berrybyrivabit]

How many fingers did you put down?

#berryinbloom #berrybyrivabit

[Nigeria Democratic Congress @NigeriaNDCHQ]

Peter Obi and Rabiu Musa Kwankwaso are here.

Medicals ongoing.

Here we go, soon!!!

[@jason @Jason]

Our government is stealing from citizens — AUDIT EVERY PROGRAM

[Akshay 🚀 @akshay_pachaar]

DeepSeek just fixed one of AI's oldest problems.

(using a 60-year-old algorithm)

Here's the story:

When deep learning took off, researchers hit a wall. You can't just stack layers endlessly. Signals either explode or vanish. Training deep networks was nearly impossible.

ResNets solved this in 2016 with residual connections:

output = input + what the layer learned

That "+" creates a direct highway for information. This is why we can now train networks with hundreds of layers.

Recently, researchers asked: what if we had multiple highways instead of one?

Hyper-Connections (HC) expanded that single lane into 4 parallel lanes with learnable matrices that mix information between streams.

The performance gains were real. But there was a problem:

Those mixing matrices compound across layers. A tiny 5% amplification per layer becomes 18x after 60 layers. The paper measured amplification reaching 3000x. Training collapses.

The usual fixes? Gradient clipping. Careful initialization. Hoping things work out.

These are hacks. And hacks don't scale.

DeepSeek went back to first principles. What mathematical constraint would guarantee stability?

The answer was sitting in a 1967 paper: the Sinkhorn-Knopp algorithm.

It forces mixing matrices to be "doubly stochastic," where rows and columns each sum to 1.

The results:

- 3000x instability reduced to 1.6x
- Stability guaranteed by math, not luck
- Only 6.7% additional training overhead

No hacks. Just math.

I've shared link to the paper in the next tweet.

[shirish @shiri_shh]

AI wrapped 2025

[Nairametrics @Nairametrics]

It’s not what you think about your idea that matters most, it’s what the market thinks.

Before you invest heavily or fall in love with it, test it. Listen. Learn. Adjust.

Ideas don’t win because they’re secret or unique. They win because the time is right and people are ready for them.

🎥myaccessbank

[Somto Okonkwo @General_Somto]

Man Calls Out Medplus Pharmacy In Nigeria: