Comments (2)
Take a look at the examples (click the little boxes below). I'm guessing the model being used has a bias for predicting positive, such that even replacing most tokens in your long review with UNK does not cause a prediction of negative. If that is the case, an empty anchor or any arbitrary anchor is 'correct', i.e. under the assumption that the perturbation is replacing things with UNK this is a sufficient condition for a prediction of 'positive'.
from anchor.
Ah okay, I understand it now, thanks!
It just surprised me that out of this long text as input, only one word was selected as the anchor. But I understand it with those examples better.
from anchor.
Related Issues (20)
- Spacy is a dependency, despite seemingly not being used or supported
- How to perturb with different format of data
- Error occuered when ran demo "Anchor on tabular data.ipynb"
- Anchor on tabular data.ipynb HOT 1
- Image-based tutorial HOT 2
- Is it possible to get empty anchors? HOT 4
- Multiple rows passed in predict function HOT 1
- Batch learning AnchorTabularExplainer? HOT 1
- Is it possible to use AnchorText with Tokenizer instead of CountVectorizer? HOT 2
- Error when using save_to_file() - IndexError: [E040] Attempt to access token at x, max length x. HOT 1
- Important features list is randomly changing HOT 1
- Anchor provided is unexpcted.
- Anchor for regression? HOT 3
- Possible anchor coverage issue
- Justification for removing bisection for computing KL-confidence regions HOT 1
- # TODO: precision recall is all wrong, coverage functions wont work
- anchor rules for time series classification HOT 2
- Get anchors from tuple method
- importing the anchor_tabular
- working with new tabular data
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from anchor.