frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Neural networks that learn non-linearity without activation functions [pdf]

https://www.tahabouhsine.com/nmn/assets/deep_learning_two_point_o_point_one.pdf
23•mlnomadpy•6mo ago

Comments

mlnomadpy•6mo ago
hello everyone, am taha,

I was able to create a new kernel that allows you to learn non-linearity without using activation functions, making the models whitebox, and without any information loss.

MiniGPT with huggingface datasets streaming: https://www.kaggle.com/code/skywolfmo/yat-nnx-minigpt-finewe...

rytill•6mo ago
Why would one have motivation to not use activation functions?

To my knowledge they’re a negligible portion of the total compute during training or inference and work well to provide non-linearity.

Very open to learning more.

russfink•6mo ago
One reason might be expressing the constructs in a different domain, eg homomorphic encrypted evaluators.
julius•6mo ago
Less information loss -> Less params? Please correct me if I got this wrong. The Intro claims:

"The dot product itself is a geometrically impoverished measure, primarily capturing alignment while conflating magnitude with direction and often obscuring more complex structural and spatial relationships [10, 11, 4, 61, 17]. Furthermore, the way current activation functions achieve non-linearity can exacerbate this issue. For instance, ReLU (f (x) = max(0, x)) maps all negative pre-activations, which can signify a spectrum of relationships from weak dissimilarity to strong anti-alignment, to a single zero output. This thresholding, while promoting sparsity, means the network treats diverse inputs as uniformly orthogonal or linearly independent for onward signal propagation. Such a coarse-graining of geometric relationships leads to a tangible loss of information regarding the degree and nature of anti-alignment or other neg- ative linear dependencies. This information loss, coupled with the inherent limitations of the dot product, highlights a fundamental challenge."

mlnomadpy•6mo ago
yes, since you can learn to represent the same problem with less amount of params, however most of the architectures are optimized for the linear product, so we gotta figure out a new architecture for it
mlnomadpy•6mo ago
they are one of the reasons neural networks are blackbox, we lose information about the data manifold the deeper we go in the network, making it impossible to trace back the output

this preprint is not coming from a standpoint of optimizing the inference/compute, but from trying to create models that we can interpret in the future and control

nikolayasdf123•6mo ago
I misread this as if "there is no non-linearity". there is still non-linearity, it is just renamed and reshuffled into new operators. basically renaming apples into oranges.
imtringued•6mo ago
Well, it's more like fruits and vegetables. The author proposed a normalized inner product as replacement for the standard inner product.

It's not an activation function, because it has the learnable weights of a linear projection (mat vec multiplication) and the clamping properties of an activation function all in one.

My personal issue with the proposal is that it essentially doubles the amount of memory needed on-chip.

Yat-Product GEMMV now needs to store the running total of the inner product and the norm of the input vectors. That's a big cost increase for something that might not improve performance all that much.

mlnomadpy•6mo ago
that's a great point you made, but the goal of this research paper is not to improve the performance, but to show that you can train deep neural networks without the need of activation functions, normalization layers, deep neural networks.

one simple usecase for them is physics-informed neural networks and neural ODEs, where using activation functions is discouraged, mainly because they aren't infinitly differentiable, and they use the tanh or the sin most of the time, this kernel i introduced works better then the neurons followed with a tanh to solve different PDEs

mlnomadpy•6mo ago
basicly the real "non-linearity" in deep learning have always been the orthogonality, squashing functions make it easy for the neurons to tap into the orthogonality, while most of the activation functions "lie" about their orthogonality by setting the dot product score to "0", and a dot product of 0 between two vectors means they are orthogonal (linear indep)

what i did was rely on both the angular information and spatial information between the input x and the weight w to measure how "similar" they are.

the lower bound of the yat-product is 0, and it is achieved only when two vectors are orthogonal and away

nurettin•6mo ago
One interesting thing to notice is how you can remodel xor into being a linear function by using u + v as input 1 and u * v as input 2 which means it can be represented in a NN without a hidden layer. And not only xor, it keeps all other logic gates simple. So only by transforming inputs one can reduce network complexity. Perhaps a field ripe for research.
mlnomadpy•6mo ago
indeed, there is an extensive work done in kernel learning that is facinating and one of the applications that still do these transformations are satellite imagery/multispectral imagery, you can get more information just from calculating the ndvi from the different bands of your image, which make it easy for your models to make decisions

We Mourn Our Craft

https://nolanlawson.com/2026/02/07/we-mourn-our-craft/
110•ColinWright•1h ago•84 comments

Speed up responses with fast mode

https://code.claude.com/docs/en/fast-mode
22•surprisetalk•1h ago•22 comments

U.S. Jobs Disappear at Fastest January Pace Since Great Recession

https://www.forbes.com/sites/mikestunson/2026/02/05/us-jobs-disappear-at-fastest-january-pace-sin...
118•alephnerd•2h ago•74 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
121•AlexeyBrin•7h ago•24 comments

Stories from 25 Years of Software Development

https://susam.net/twenty-five-years-of-computing.html
62•vinhnx•5h ago•7 comments

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

https://openciv3.org/
827•klaussilveira•21h ago•248 comments

Al Lowe on model trains, funny deaths and working with Disney

https://spillhistorie.no/2026/02/06/interview-with-sierra-veteran-al-lowe/
55•thelok•3h ago•7 comments

Brookhaven Lab's RHIC Concludes 25-Year Run with Final Collisions

https://www.hpcwire.com/off-the-wire/brookhaven-labs-rhic-concludes-25-year-run-with-final-collis...
4•gnufx•37m ago•0 comments

The AI boom is causing shortages everywhere else

https://www.washingtonpost.com/technology/2026/02/07/ai-spending-economy-shortages/
108•1vuio0pswjnm7•8h ago•136 comments

The Waymo World Model

https://waymo.com/blog/2026/02/the-waymo-world-model-a-new-frontier-for-autonomous-driving-simula...
1058•xnx•1d ago•611 comments

Reinforcement Learning from Human Feedback

https://rlhfbook.com/
76•onurkanbkrc•6h ago•5 comments

I Write Games in C (yes, C)

https://jonathanwhiting.com/writing/blog/games_in_c/
8•valyala•1h ago•1 comments

Start all of your commands with a comma (2009)

https://rhodesmill.org/brandon/2009/commands-with-comma/
484•theblazehen•2d ago•175 comments

SectorC: A C Compiler in 512 bytes

https://xorvoid.com/sectorc.html
7•valyala•2h ago•0 comments

Vocal Guide – belt sing without killing yourself

https://jesperordrup.github.io/vocal-guide/
209•jesperordrup•12h ago•70 comments

France's homegrown open source online office suite

https://github.com/suitenumerique
557•nar001•6h ago•256 comments

Coding agents have replaced every framework I used

https://blog.alaindichiappari.dev/p/software-engineering-is-back
222•alainrk•6h ago•343 comments

A Fresh Look at IBM 3270 Information Display System

https://www.rs-online.com/designspark/a-fresh-look-at-ibm-3270-information-display-system
36•rbanffy•4d ago•7 comments

Selection Rather Than Prediction

https://voratiq.com/blog/selection-rather-than-prediction/
8•languid-photic•3d ago•1 comments

History and Timeline of the Proco Rat Pedal (2021)

https://web.archive.org/web/20211030011207/https://thejhsshow.com/articles/history-and-timeline-o...
19•brudgers•5d ago•4 comments

72M Points of Interest

https://tech.marksblogg.com/overture-places-pois.html
29•marklit•5d ago•2 comments

Unseen Footage of Atari Battlezone Arcade Cabinet Production

https://arcadeblogger.com/2026/02/02/unseen-footage-of-atari-battlezone-cabinet-production/
114•videotopia•4d ago•31 comments

Show HN: I saw this cool navigation reveal, so I made a simple HTML+CSS version

https://github.com/Momciloo/fun-with-clip-path
5•momciloo•1h ago•0 comments

Where did all the starships go?

https://www.datawrapper.de/blog/science-fiction-decline
76•speckx•4d ago•75 comments

Show HN: Look Ma, No Linux: Shell, App Installer, Vi, Cc on ESP32-S3 / BreezyBox

https://github.com/valdanylchuk/breezydemo
273•isitcontent•22h ago•38 comments

Show HN: Kappal – CLI to Run Docker Compose YML on Kubernetes for Local Dev

https://github.com/sandys/kappal
22•sandGorgon•2d ago•11 comments

Learning from context is harder than we thought

https://hy.tencent.com/research/100025?langVersion=en
201•limoce•4d ago•111 comments

Monty: A minimal, secure Python interpreter written in Rust for use by AI

https://github.com/pydantic/monty
286•dmpetrov•22h ago•153 comments

Making geo joins faster with H3 indexes

https://floedb.ai/blog/how-we-made-geo-joins-400-faster-with-h3-indexes
155•matheusalmeida•2d ago•48 comments

Software factories and the agentic moment

https://factory.strongdm.ai/
71•mellosouls•4h ago•75 comments