Der smart Concept #5 ist das bisher geräumigste und vielseitigste Fahrzeug von smart und markiert den Einstieg der rein elektrischen Marke in das Premium-SUV-Segment der Mittelklasse.

smart einfach ganz anders: smart Concept #5 - oekonews.at

Smartz

The number of public companies that are taking positions in bitcoin is simply staggering. 

From Fiat Money to Ethical DeFi 

<strong>Cryptocurrency tokens taking the market by storm </strong><br>Cryptocurrency tokens are marching into the market with quite a pace now. However, not everybody has a clear vision of what exactly do tokens represent and how they function. For this, we would like to give a little bit of introduction on ERC20 (arguably the most popular type of cryptocurrency tokens) as well as mention several services that might help you deploy your own token.

How to create an ERC20 token?

With a good algorithm, you can simultaneously solve several problems associated with smart contracts that work with large lists of user addresses. You can’t add a list of several thousand addresses to a contract and allow this set of addresses to do something in that contract. The <a href="https://hackernoon.com/tagged/blockchain" target="_blank">blockchain</a> tries to save every single byte, so it would be too expensive to have so much data there.

Evolution of Airdrop: from Common Spam to the Merkle Tree

Why do smart teams choose known failure over uncertain change? Explore the cognitive glitches—like the Ambiguity Effect—shaping the modern digital labyrinth. 

The Architecture of Stagnation - Why Smart Tech Teams Prefer 'Known Failure' Over the Unknown

A new cyber-insurance model uses smart monitoring, epidemic risk modeling, and Shapley values to fairly price cyber risk in power grids.

A New Cyber-Insurance Model Uses Game Theory to Protect Smart Power Grids

OptikosPrime just obliterated the primary barrier in digital optometry. In a world first, its Argus platform provides vision tests based on a selfie.

OptikosPrime Argus Poised to Transform Industry with First in Smartphone Vision Tests

Examine Motorola’s pivot to IoT security and smart home surveillance. We highlight key features and infrastructure to decode what's important in home safety. 

Beyond Smartphones: Motorola's Pivot to IoT Security and Smart Home Surveillance

The “misalignment window” reveals whether an AI agent feels sharp or strangely distant.

The One Latency Metric That Tells You If Your AI Agent Is Actually Smart

The Hidden Reason Smart Homes Keep Making Dumb Energy Decisions

<p><em>This is a Plain English Papers summary of a research paper called <a href="https://www.aimodels.fyi/papers/arxiv/mhc-manifold-constrained-hyper-connections?utm_source=hackernoon&utm_medium=referral">mHC: Manifold-Constrained Hyper-Connections</a>. If you like these kinds of analysis, join <a href="https://www.aimodels.fyi/?utm_source=hackernoon&utm_medium=referral">AIModels.fyi</a> or follow us on <a href="https://x.com/aimodelsfyi">Twitter</a>.</em></p>

<h2>The hidden genius behind residual connections</h2>
<p>Residual connections changed deep learning fundamentally. The idea is simple: instead of each layer processing information fresh, you add the original input back to the output. So y equals f(x) plus x. This seemingly small change unlocked the ability to train networks with hundreds of layers without everything falling apart during training.</p>
<p>The reason this works comes down to gradient flow. When you train a neural network, you calculate gradients that tell you how to adjust each parameter. In deep networks without residual connections, these gradients either vanish to nothing or explode to infinity as they propagate backward through many layers. The identity mapping created by the residual connection gives gradients a direct highway to travel back through the network unchanged. Early layers still receive meaningful learning signals even in very deep networks.</p>
<p>This property made residual connections so fundamental that every major architecture built in the last decade relies on them, from Transformers to modern language models. What started as an architectural trick became a foundational principle.</p>
<h2>Why wider connections seemed like an easy win</h2>
<p>Researchers naturally asked: if one residual bypass works well, what if you created multiple bypasses with different paths? This is the idea behind Hyper-Connections, which expand the residual stream width and diversify connectivity patterns. Instead of a single connection between layers, you'd have richer networks of information flowing in parallel. The intuition seemed sound, and early work showed real performance improvements.</p>
<p>But this expansion came with a hidden cost. When you add multiple pathways and widen the connection space, you fundamentally change how the connections work. The function combining those paths no longer preserves the identity mapping property. You've gained architectural complexity but lost the mathematical guarantee that made residual connections stable in the first place.</p>
<p>This loss of the identity mapping created two serious problems. First, training became unstable. Gradients behaved erratically during backpropagation, making it difficult to scale these networks to realistic sizes. Second, moving data through those wider connections consumed substantial memory, creating computational overhead that eroded the practical benefits. The performance gains came at a cost that grew with scale.</p>
<p>Related work on <a href="https://aimodels.fyi/papers/arxiv/hyper-connections?utm_source=hackernoon&utm_medium=referral">Hyper-Connections</a> and <a href="https://aimodels.fyi/papers/arxiv/frac-connections-fractional-extension-hyper-connections?utm_source=hackernoon&utm_medium=referral">Fractional extensions</a> had explored these wider connection patterns, but neither addressed the fundamental flaw: the loss of the stability property that makes residual connections work.</p>
<h2>The mathematical constraint that fixes everything</h2>
<p>This is where the paper reveals its core insight. You don't have to choose between architectural complexity and training stability. Instead, you constrain where that complexity lives.</p>
<p>Think of a sphere. You can move in many directions on its surface, but you're always constrained to the spherical structure itself. You haven't lost freedom, you've shaped it. The paper applies this same logic to neural network connections: allow rich, diverse hyper-connections, but only if they live on a specific manifold, a lower-dimensional mathematical surface embedded in the high-dimensional connection space.</p>
<p>The key is that this manifold constraint preserves the identity mapping property locally. Even though the connections are wider and more complex, the way they combine respects the fundamental principle that makes residual connections work. The hyper-connections get projected onto a manifold that includes the identity function itself. This isn't a compromise that trades away performance. It's a structural constraint that allows you to have both complexity and stability.</p>
<p>The mathematical elegance matters because it resolves the tension completely. You get the stability of the original residual connection design with the performance potential of the wider architecture. Training behaves properly because gradients flow through paths that respect the identity mapping property. The manifold acts as guardrails, keeping you in a learnable zone while still exploring the expanded architectural space.</p>
<h2>Making it actually efficient</h2>
<p>Mathematics that doesn't run efficiently is rarely useful. The paper doesn't stop at theory, it includes infrastructure optimizations that exploit the manifold structure to reduce memory overhead and computational cost.</p>
<p>Adding mathematical structure to a problem often enables more efficient computation. The manifold constraint provides this structure naturally. Instead of shuffling data through arbitrarily wide connections, the manifold structure allows more efficient implementations. The result is both better performance and better efficiency, which rarely coexist without engineering compromise.</p>
<p>This matters because it separates mHC from purely theoretical contributions. The constraint isn't a beautiful idea that only works on toy problems. It's something you could actually use when training real models with billions of parameters. The optimization work shows that the theoretical insight translates into practical advantage.</p>
<h2>Testing at real scale</h2>
<p>The paper's claims need evidence. Do the theoretical benefits actually materialize when training realistic models? The experiments test three specific questions: Does mHC maintain the performance improvements of Hyper-Connections while fixing the training instability? Does it actually scale to large models without the memory and computational overhead that plagued standard HC? How does it compare to both baseline residual connections and the wider hyper-connections it improves upon?</p>
<p>The experimental results show that mHC handles the complexity trade-off gracefully. Performance doesn't drop compared to HC, meaning you're not sacrificing the gains that motivated hyper-connections in the first place. Training curves show substantially smoother learning dynamics without the instability that made HC difficult to scale. Scalability improves genuinely, allowing larger models to train with the same computational resources.</p>
<p>These results matter because they validate the entire contribution. If mHC worked only on small networks or toy problems, it would be academically interesting but practically limited. The fact that it works at real scale demonstrates that the theoretical insight translates into something useful for the architectures that power modern AI.</p>
<h2>What this means for building better models</h2>
<p>The paper solves a specific technical problem, but the implications extend further. It reveals something important about how neural network architectures actually work. Residual connections succeeded not because they're the only way to build networks, but because they preserve a specific mathematical property while adding functionality. When you try to extend that design, you risk losing that property unless you're strategic about it.</p>
<p>This points toward a broader principle in topological architecture design, the study of how information flows through network structure. Rather than simply trying new architectures and seeing what works, you can understand the underlying principles that make architectures successful, then innovate within constraints that preserve those principles. It's the difference between trial and error and principled design.</p>
<p>The work on <a href="https://aimodels.fyi/papers/arxiv/deep-manifold-part-2-neural-network-mathematics?utm_source=hackernoon&utm_medium=referral">deep manifolds</a> and network mathematics suggests this approach scales to other architectural decisions. The lesson applies broadly: preservation and innovation coexist if you find the right constraints.</p>
<p>For foundational models, the giant networks that power modern AI systems, this matters deeply. These models are built on architectural principles refined over years of research. If you understand how to innovate responsibly, preserving the properties that make things work while adding new capability, you can guide the evolution of these models more effectively. You move from architecture as empirical craft toward architecture as principled design, where changes are motivated by understanding rather than just intuition.</p>
<p>The paper's real contribution isn't any single technical detail. It's the recognition that you don't need to choose between preserving a foundational principle and innovating beyond it. You can do both when you find the right mathematical structure. That insight will likely shape how future architectures develop.</p>

<hr/><p><strong>Original post:</strong> <a href="https://www.aimodels.fyi/papers/arxiv/mhc-manifold-constrained-hyper-connections?utm_source=hackernoon&utm_medium=referral">Read on AIModels.fyi</a></p>

The Math Trick That Lets Deep Networks Get Smarter Without Falling Apart

Are those toys secure? And precisely what data is being handed over when a kid is using these toys?

How You Can Test Your Kids' Smart Toys For Privacy 

AI makes it easy to sound smart without understanding. When confidence is rewarded over correctness, systems quietly fail.

Why Sounding Smart Is No Longer Proof of Knowing

We are a PropTech SaaS company whose mission is to make the lives of property owners and tenants easier with our all-in-one software

SMARTZ

smartz.cloud

33 emps

Since 2021

SMARTZ

EVERGREEN INDEX #4033

Smartz's stories on HackerNoon

From Fiat Money to Ethical DeFi

How to create an ERC20 token?

Evolution of Airdrop: from Common Spam to the Merkle Tree

The Architecture of Stagnation - Why Smart Tech Teams Prefer 'Known Failure' Over the Unknown

A New Cyber-Insurance Model Uses Game Theory to Protect Smart Power Grids

OptikosPrime Argus Poised to Transform Industry with First in Smartphone Vision Tests

Beyond Smartphones: Motorola's Pivot to IoT Security and Smart Home Surveillance

The One Latency Metric That Tells You If Your AI Agent Is Actually Smart

The Hidden Reason Smart Homes Keep Making Dumb Energy Decisions

The Math Trick That Lets Deep Networks Get Smarter Without Falling Apart

How You Can Test Your Kids' Smart Toys For Privacy

Why Sounding Smart Is No Longer Proof of Knowing

Smartz's latest news & mentions

smart einfach ganz anders: smart Concept #5 - oekonews.at

Smartz's Related Companies

Representing this company? Click here to claim and customize this page!

Reviews About Smartz

Smartz WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps

SMARTZ

smartz.cloud

33 emps

Since 2021

SMARTZ

EVERGREEN INDEX #4033

Smartz's stories on HackerNoon

From Fiat Money to Ethical DeFi

How to create an ERC20 token?

Evolution of Airdrop: from Common Spam to the Merkle Tree

The Architecture of Stagnation - Why Smart Tech Teams Prefer 'Known Failure' Over the Unknown

A New Cyber-Insurance Model Uses Game Theory to Protect Smart Power Grids

OptikosPrime Argus Poised to Transform Industry with First in Smartphone Vision Tests

Beyond Smartphones: Motorola's Pivot to IoT Security and Smart Home Surveillance

The One Latency Metric That Tells You If Your AI Agent Is Actually Smart

The Hidden Reason Smart Homes Keep Making Dumb Energy Decisions

The Math Trick That Lets Deep Networks Get Smarter Without Falling Apart

How You Can Test Your Kids' Smart Toys For Privacy

Why Sounding Smart Is No Longer Proof of Knowing

Smartz's latest news & mentions

smart einfach ganz anders: smart Concept #5 - oekonews.at

Smartz's Related Companies

2015

9.3B

-1%

2014

7.9B

2015

5B

2015

4.8B

2%

2012

4.3B

-1%

2013

4B

Representing this company? Click here to claim and customize this page!

Reviews About Smartz

Smartz WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps