Rendered at 12:19:54 GMT+0000 (Coordinated Universal Time) with Vercel.
folderquestion 1 days ago [-]
This sound like projecting data into the linear space spanned by {x_i, x_i*x_j} where x_i are the features variables, and then applying standard regularization methods to remove noise and low value coefficients.
Anisotropy and the cone ideas may explain why PCA underperforms, but it does not uniquely justify this particular quadratic decoder. The geometric story is not doing explanatory work beyond “data is nonlinear,” and the real substance is simply that second-order reconstruction empirically helps.
wenc 21 hours ago [-]
It sounds like this replaces the PCA reconstruction function with a quadratic.
The normal PCA encoding:
1) Given a mean-center-scaled X matrix, get the latent variable matrix T with X = T * P’ + e, where P = loadings and e = residuals. The P is your model, so for a new vector xnew, you can calculate tnew = xnew * P (because P’ * P = I).
This is the encoder —- nothing changes here. The original matrix is dimensionally reduced with residuals e discarded. This is why PCA is lossy.
The decoder is where things diverge
The usual PCA decoder reconstructs a given latent variable t_any by using the trained P loadings, like thus x_reconstructed = t_any * P’. This reconstructed data lies on a linear hyperplane, so if the original data did not lie on the hyperplane, reconstruction errors are potetially high.
In your proposal, instead of a linear decoder, you train a quadratic decoder (essentially a classic ridge regression using a quadratic) on the original X. So for your reconstruction, you have x_reconstructed = poly(t_new).
This achieves lower reconstruction error in-sample (naturally, because quadratic is higher order than linear), but your poly function is trained on a particular corpus. Which means that when you’re in-distribution within that corpus, you’re good but when you’re not, you can be very wrong in biased ways that PCA’s linear reconstruction is not.
SO this is not a better technique than PCA in a general sense. It’s a better reconstruction machine when your data is mostly in-sample. It’s a kind of computationally cheap “specialization” on a particular distribution of data, which can be useful if you’re mostly in-distribution but introduces new risks when out-of-distribution.
Whereas PCA just drops the residual and makes modest claims, a quadratic decoder is trying to predict the residual and on out-of-sample data, it can be wrong in biased ways that PCA is not. In other words, it can hallucinate.
But if on a large enough training corpus, chances are we’re going to be in-distribution most of the time, so maybe this could generalize well.
20 hours ago [-]
rgovostes 19 hours ago [-]
This article uses every one of Claude's cliches, e.g., "No SGD, no epochs, no hyperparameter search." It's hard to tell if this is real research.
ComplexSystems 4 hours ago [-]
What makes this different from just kernel PCA with the quadratic kernel?
pleshkov 4 days ago [-]
Author here — questions and pushback both welcome.
Devilstro 1 days ago [-]
In the article, you mention this approach requires no search over hyper-parameter, because the method comprises a closed-form solution with "simple" linear algebra. I agree with this, but do you not in think need to tune the L2-regularization strength? That would for me be a hyper-parameter you would need to do a CV over (or similarly).
yorwba 1 days ago [-]
You should benchmark the retrieval speed of each method in terms of queries per second. I suspect that the gain in bandwidth you get from slightly better compression will be defeated by decompression being much more expensive.
afxuh 1 days ago [-]
Cool idea. But it only works when the data never changes.
could you make a streaming/incremental version? One that updates the math cheaply when new data arrives, instead of recomputing everything, or does the math fundamentally prevent it?
mpaiello 1 days ago [-]
[dead]
stephantul 1 days ago [-]
Really cool! I was investigating PCA on retrieval, thanks for the references.
yobbo 1 days ago [-]
My understanding after scanning the code examples is the technique expands the dimensionality of each data point with a set consisting of the quadratic coefficients of its existing dimensions. I thought it sounded like kernel PCA.
ekjhgkejhgk 1 days ago [-]
I think that kernel PCA is a strict superset of PCA. That would make it trivially true that it beats PCA.
mentalgear 1 days ago [-]
Geometric Algebra (GA) (Clifford Algebra) also has high potential to transform neural architectures. Models like the Geometric Algebra Transformer (GATr) and Versor (2026) demonstrate it can enhance or even make the Attention Mechanism obsolete.
By representing data as multivectors, translational and rotational symmetries are encoded natively which allows them to handle geometric hierarchies with massive efficiency gains (reports of up to 78x speedups and 200x parameter reductions) compared to standard Transformers.
> A novel sequence architecture is introduced, Versor, which uses Conformal Geometric Algebra (CGA) in place of traditional linear operations to achieve structural generalization and significant performance improvements on a variety of tasks, while offering improved interpretability and efficiency. By embedding states in the manifold and evolving them via geometric transformations (rotors), Versor natively represents -equivariant relationships without requiring explicit structural encoding. Versor is validated on chaotic N-body dynamics, topological reasoning, and standard multimodal benchmarks (CIFAR-10, WikiText-103), consistently outperforming Transformers, Graph Networks, and geometric baselines (GATr, EGNN).
I'm just a casual LLM user, but your description of the anisotropy made me think about the recent work on KV cache quantization techniques such as TurboQuant where they apply a random rotation on each vector before quantizing, as I understood it precisely to make it more isotropic.
But for RAG that might be too much work per vector?
electroglyph 1 days ago [-]
this looks awesome. i've been struggling with vector compression, and have been trying PCA + all sorts of rotations. looking forward to trying this out
Abhijeet620380 17 hours ago [-]
Nice idea. The UI feels simple and easy to use.
nickpsecurity 20 hours ago [-]
There's also work in polynomial regression as an alternative to regular, neural networks.
I came here from a discussion about CS students who should not be bothered to set up email filters. How can they ever expect to be able to digest just the first paragraph in that article?
whywhywhywhy 1 days ago [-]
This kind of attitude isn’t long for this world. The time for this knowledge locked up in academia is over and in 15 years we’ll look back on this time as the dark ages as open code and models eclipses it.
None of this stuff is as difficult to understand as people claim it is once you work with it.
magicalhippo 1 days ago [-]
FWIW I found it quite straight forward. But then I did have some linear algebra back at uni.
That said I do think it's a good habit to either write out abbreviations in full or link to say Wikipedia, eg for PCA[1]. It's a well-known tool but still if you come from a slightly different field it might not ring a bell.
Anisotropy and the cone ideas may explain why PCA underperforms, but it does not uniquely justify this particular quadratic decoder. The geometric story is not doing explanatory work beyond “data is nonlinear,” and the real substance is simply that second-order reconstruction empirically helps.
The normal PCA encoding:
1) Given a mean-center-scaled X matrix, get the latent variable matrix T with X = T * P’ + e, where P = loadings and e = residuals. The P is your model, so for a new vector xnew, you can calculate tnew = xnew * P (because P’ * P = I).
This is the encoder —- nothing changes here. The original matrix is dimensionally reduced with residuals e discarded. This is why PCA is lossy.
The decoder is where things diverge
The usual PCA decoder reconstructs a given latent variable t_any by using the trained P loadings, like thus x_reconstructed = t_any * P’. This reconstructed data lies on a linear hyperplane, so if the original data did not lie on the hyperplane, reconstruction errors are potetially high.
In your proposal, instead of a linear decoder, you train a quadratic decoder (essentially a classic ridge regression using a quadratic) on the original X. So for your reconstruction, you have x_reconstructed = poly(t_new).
This achieves lower reconstruction error in-sample (naturally, because quadratic is higher order than linear), but your poly function is trained on a particular corpus. Which means that when you’re in-distribution within that corpus, you’re good but when you’re not, you can be very wrong in biased ways that PCA’s linear reconstruction is not.
SO this is not a better technique than PCA in a general sense. It’s a better reconstruction machine when your data is mostly in-sample. It’s a kind of computationally cheap “specialization” on a particular distribution of data, which can be useful if you’re mostly in-distribution but introduces new risks when out-of-distribution.
Whereas PCA just drops the residual and makes modest claims, a quadratic decoder is trying to predict the residual and on out-of-sample data, it can be wrong in biased ways that PCA is not. In other words, it can hallucinate.
But if on a large enough training corpus, chances are we’re going to be in-distribution most of the time, so maybe this could generalize well.
By representing data as multivectors, translational and rotational symmetries are encoded natively which allows them to handle geometric hierarchies with massive efficiency gains (reports of up to 78x speedups and 200x parameter reductions) compared to standard Transformers.
> A novel sequence architecture is introduced, Versor, which uses Conformal Geometric Algebra (CGA) in place of traditional linear operations to achieve structural generalization and significant performance improvements on a variety of tasks, while offering improved interpretability and efficiency. By embedding states in the manifold and evolving them via geometric transformations (rotors), Versor natively represents -equivariant relationships without requiring explicit structural encoding. Versor is validated on chaotic N-body dynamics, topological reasoning, and standard multimodal benchmarks (CIFAR-10, WikiText-103), consistently outperforming Transformers, Graph Networks, and geometric baselines (GATr, EGNN).
https://arxiv.org/abs/2602.10195
But for RAG that might be too much work per vector?
Polynomial Regression As an Alternative to Neural Nets (2018) https://arxiv.org/abs/1806.06850
Π-nets: Deep Polynomial Neural Networks (2020) https://arxiv.org/abs/2003.03828
None of this stuff is as difficult to understand as people claim it is once you work with it.
That said I do think it's a good habit to either write out abbreviations in full or link to say Wikipedia, eg for PCA[1]. It's a well-known tool but still if you come from a slightly different field it might not ring a bell.
[1]: https://en.wikipedia.org/wiki/Principal_component_analysis