AI News HubLIVE
Original source2 min read

Hierarchical Grading in Large Language Models

This paper introduces Graded Large Language Models (GLLMs), an algebraic framework that equips the transformer representation space with a grading and propagates the induced weighted scalar action through embeddings, self-attention, and the training objective. It extends graded neural networks and graded transformers to autoregressive language models while preserving expressive power and complexity. The geometric picture uses invariant theory, where the benefits of grading are expressed via a Kempf–Ness functional, and optimal grades come from a coincidence point of two moment maps. A standard transformer appears as a semistable isotropic point on the boundary. For level-stratified targets, a minimax separation is proven between graded and uniform priors, with an exponential decay in risk.

SourcearXiv Machine LearningAuthor: T. Shaska

-->

[Submitted on 23 Jul 2026]

Title:Hierarchical Grading in Large Language Models

View a PDF of the paper titled Hierarchical Grading in Large Language Models, by T. Shaska

View PDF HTML (experimental)

Abstract:We introduce Graded Large Language Models (GLLMs), an algebraic framework that equips the representation space of a transformer with a grading and propagates the induced weighted scalar action through embeddings, self-attention, and the training objective. The construction extends the theory of graded neural networks and graded transformers to autoregressive language models while preserving expressive power, asymptotic computational complexity, and inference cost.

The governing geometric picture is that of geometric invariant theory. The benefit of a grading is expressed by a Kempf--Ness functional on the grading torus; the grades that improve upon the uniform architecture form an open convex cone whose membership is decided by a Hilbert--Mumford-type criterion pairing a grade direction against two measurable profiles of the target and the data; the optimal grades are the coincidence point of two moment maps, given in closed form; and the ordinary transformer appears as a semistable isotropic point on the boundary of the cone: one member of a larger graded family rather than a distinguished optimum.

Separately, for level-stratified targets we prove a minimax separation between the graded prior and its absence: over all estimators the risks of the graded and uniform target classes separate throughout an explicit window of sample sizes, by a factor that decays exponentially in the number of levels under geometric stratification. Both profiles are estimable offline, so the optimal grades solve a convex program certified before training begins. Because the grading is absorbed into the learned parameters after training, every GLLM compiles to a standard transformer of identical architecture and inference complexity.

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI)

ACM classes: I.2.7; I.2.6; F.2.2; H.1.1

Cite as: arXiv:2607.22757 [cs.LG]

(or arXiv:2607.22757v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2607.22757

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Tanush Shaska [view email] [v1] Thu, 23 Jul 2026 20:14:48 UTC (65 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Hierarchical Grading in Large Language Models, by T. Shaska

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.LG

new | recent | 2026-07

Change to browse by:

cs cs.AI

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

IArxiv recommender toggle

IArxiv Recommender (What is IArxiv?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)