Skip to main content
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arXiv logo
Cornell University Logo

Computer Science > Computation and Language

arXiv:1809.03044 (cs)
[Submitted on 9 Sep 2018]

Title:How clever is the FiLM model, and how clever can it be?

Authors:Alexander Kuhnle, Huiyuan Xie, Ann Copestake
View PDF
Abstract:The FiLM model achieves close-to-perfect performance on the diagnostic CLEVR dataset and is distinguished from other such models by having a comparatively simple and easily transferable architecture. In this paper, we investigate in more detail the ability of FiLM to learn various linguistic constructions. Our main results show that (a) FiLM is not able to learn relational statements straight away except for very simple instances, (b) training on a broader set of instances as well as pretraining on simpler instance types can help alleviate these learning difficulties, (c) mixing is less robust than pretraining and very sensitive to the compositional structure of the dataset. Overall, our results suggest that the approach of big all-encompassing datasets and the paradigm of "the effectiveness of data" may have fundamental limitations.
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
Cite as: arXiv:1809.03044 [cs.CL]
  (or arXiv:1809.03044v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.1809.03044

Submission history

From: Alexander Kuhnle [view email]
[v1] Sun, 9 Sep 2018 21:08:57 UTC (55 KB)
Full-text links:

Access Paper:

  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs.CL
< prev   |   next >
new | recent | 2018-09
Change to browse by:
cs
cs.AI
cs.CV
cs.LG

References & Citations

DBLP - CS Bibliography

listing | bibtex
Alexander Kuhnle
Huiyuan Xie
Ann A. Copestake
export BibTeX citation

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)

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?)