# limix-ldm-ai/LimiX

LimiX: Unleashing Structured-Data Modeling Capability for Generalist Intelligence https://arxiv.org/abs/2509.03505

Repository: https://github.com/limix-ldm-ai/LimiX
Canonical: https://ross.abutalabs.com/products/limix
Homepage: https://www.limix.ai
Language: Python
License: Apache-2.0
License Family: permissive
Topics: foundation-models, limix, machine-learning, structured-data
Last push: 2026-06-16T01:49:54+00:00

## Health v2 (maintenance only)
Score: 60/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 87, release rhythm 44, longevity 26
- inputs: {"age_days": 371, "days_push": 79, "days_rel": 296, "gap_med": 51, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4045, forks 305 (observed 2026-08-28T04:08:33.568057+00:00)

## What it is
LimiX is the first large structured-data foundation model (LDM), a transformer-based model for tabular data that handles classification, regression, missing-value imputation, and tabular generation with a single pretrained model. It is open-sourced under Apache 2.0 and claims to outperform XGBoost and other tabular deep learning and foundation models on mainstream structured-data benchmarks.

## Use cases
- classify tabular data without training a task-specific model
- predict continuous targets on structured datasets
- impute missing values in tabular datasets
- generate synthetic tabular data
- replace XGBoost or bespoke tabular ML pipelines with a foundation model
- perform feature selection and causal inference on structured data

## When to choose
- you need strong out-of-the-box performance on tabular classification or regression
- your datasets have missing values and you want joint modeling without separate imputation pipelines
- you want a single pretrained model covering multiple tabular tasks
- you want a zero-training or few-shot alternative to gradient boosting

## When to avoid
- you need a lightweight CPU-only solution, since inference requires GPU resources
- you work with unstructured data like text, images, or audio
- you need full interpretability of a simple model like linear regression or decision trees
- your dataset is enormous and training a specialized model is feasible and preferable

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, data-science
- domain: machine-learning, data-science, artificial-intelligence
- platform: python, cross-platform
- tags: tabular-data, foundation-model, transformer, classification, regression, missing-value-imputation, tabular-generation, xgboost-alternative, gpu

## Member repositories
- limix-ldm-ai/LimiX (main) score 60

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:33.568057+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:23:43.958735+00:00, confidence not recorded.
  - readme: https://github.com/limix-ldm-ai/LimiX (fetched 2026-08-28T04:08:33.568057+00:00, sha ade00ccbda1a)
  - homepage: https://www.limix.ai (fetched 2026-08-29T09:16:09.724695+00:00, sha 26e0968e9dd4)
- Data as of 2026-08-30T08:39:29.467469+00:00.