Neuro-Symbolic Concept Learner

Closed weights Massachusetts Institute of Technology (MIT),Tsinghua University,MIT-IBM Watson AI Lab,DeepMind April 2019

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Massachusetts Institute of Technology (MIT),Tsinghua University,MIT-IBM Watson AI Lab,DeepMind
Organisation type
Academia,Academia,Academia,Industry,Industry
Country
United States of America, China, United Kingdom of Great Britain and Northern Ireland
Published
26 April 2019
Authors
Jiayuan Mao, Chuang Gan, Pushmeet Kohli, Joshua B. Tenenbaum, Jiajun Wu

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Vision, Language
Task
Visual question answering, Semantic segmentation

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Training data
100,000 tokens

CLEVR, ImageNet, VQS 5000 in CLEVR 64509 in VQS and whole ImageNet for pretraining "We train NS-CL on 5K images (<10% of CLEVR’s 70K training images). We generate 20 questions for each image for the entire curriculum learning process" section 4.3 "All models use a pre-trained semantic parser on the full CLEVR dataset" "The only extra supervision of the visual perception module comes from the pre-training of the perception modules on ImageNet (Deng et al., 2009). To quantify the influence of th…

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased
Training code
Open source

MIT code: https://github.com/vacancy/NSCL-PyTorch-Release

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Why it is tracked
SOTA improvement

"NS-CL’s modularized design enables interpretable, robust, and accurate visual reasoning: it achieves state-of-the-art performance on the CLEVR datase"

Record confidence
Unknown
Citations
837

Sources

Where this record came from and when it was last checked.

Reference
The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision
Last updated
25 May 2026

What the numbers mean

About this model

Neuro-Symbolic Concept Learner was published by Massachusetts Institute of Technology (MIT),Tsinghua University,MIT-IBM Watson AI Lab,DeepMind, in United States of America, in April 2019. academia,Academia,Academia,Industry,Industry is the category the publisher falls under.

It works in Vision, Language, and is recorded as doing visual question answering, Semantic segmentation.

Its weights were never published, so it can only be reached through its provider. No graphics card changes that.

How it was trained

The training set ran to roughly 100,000 tokens.

It is tracked in the underlying dataset for one reason in particular: sOTA improvement.

Answers

Neuro-Symbolic Concept Learner — common questions

01

Is Neuro-Symbolic Concept Learner open source?

No. Neuro-Symbolic Concept Learner has not had its weights published, so it exists only as a service controlled by its owner.

02

How many parameters does Neuro-Symbolic Concept Learner have?

No parameter count has been published for Neuro-Symbolic Concept Learner, which is why no memory or speed figure appears on this page.

03

Who created Neuro-Symbolic Concept Learner?

Neuro-Symbolic Concept Learner was published by Massachusetts Institute of Technology (MIT),Tsinghua University,MIT-IBM Watson AI Lab,DeepMind, based in United States of America, categorised as academia,Academia,Academia,Industry,Industry.

04

When was Neuro-Symbolic Concept Learner released?

Neuro-Symbolic Concept Learner was published in April 2019. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

05

What is Neuro-Symbolic Concept Learner used for?

Neuro-Symbolic Concept Learner works in Vision, Language, and is recorded as handling visual question answering, Semantic segmentation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

06

What GPU do I need to run Neuro-Symbolic Concept Learner?

None. Neuro-Symbolic Concept Learner is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

Source

Original publication

Record last updated 25 May 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.