Zining's Blog

Home Archives About
2022-07-31
Academia

OOD-Probe: A Neural Interpretation of Out-of-Domain Generalization

This paper probes the behaviors of DNNs in generalizations. We find that the information of out-of-domain generalization encodes more “linearly” on the intermediate layers than other layers for VLCS and PACS, but not so for RotatedMNIST.

Read More

Share Comments
  • ExplainableAI
  • OOD
2022-07-30
Academia

Out-of-Distribution Failure through the Lens of Labeling Mechanisms: An Information Theoretic Approach

Introducing additional labels can help OOD generalization performance.

Read More

Share Comments
  • ExplainableAI
  • OOD
2022-04-23
Academia

On the data requirements of probing

This paper uses learning theory to recommend suitable dataset sizes to reliably reproduce probing findings.

Read More

Share Comments
  • ExplainableAI
  • NLP
2022-04-23
Academia

Neural reality of argument structure constructions

This paper adopts psycholinguistic studies on neural language models to analyze where the meanings are carried.

Read More

Share Comments
  • NLP
  • linguistics
2021-12-28
Literature

Cool Books in 2021

Here are some books I found interesting in 2021.

  1. 西方现代思想讲义(刘擎)
  2. Capital and Ideology (Piketty)
  3. Keep Sharp: Build a Better Brain at Any Age (Gupta)
  4. The Nature of Expertise (Chi etal)

Read More

Share Comments
  • book-review
2021-11-07
Academia

Quantifying the Task-Specific Information in Text-Based Classifications

In this paper, we estimate the task-specific information in text-based classification datasets.

Read More

Share Comments
  • ExplainableAI
  • Information
  • NLP
2021-11-07
Academia

An unsupervised framework for tracing textual sources of moral change

In this paper, we propose an unsupervised framework that infers the source text with prominent influence on the moral time course.

Read More

Share Comments
  • ExplainableAI
  • NLP
  • Society
2021-07-20
Academia

What do writing features tell us about AI papers?

This paper studies the writing features and their indications towards conference / workshop appearances in AI venues.

Read More

Share Comments
  • ExplainableAI
  • NLP
2021-07-10
Academia

How is BERT surprised? Layerwise detection of linguistic anomalies

This paper uses Gaussian mixture model to detect how contextualized LMs show surprisal given different types of anomalies.

Read More

Share Comments
  • ExplainableAI
  • NLP
  • linguistics
2021-06-10
Academia

NAACL 2021 papers

NAACL 2021 shows papers on the Underline.io platform. Different from miniconf (which hosted ACL 2020), Underline has a slightly different UI. I was able to go through the videos and take screenshots of a lot of posters in the notes.

Read More

Share Comments
  • NLP
  • conferences
123…5Next »

Recent Posts

  • • OOD-Probe: A Neural Interpretation of Out-of-Domain Generalization
  • • Out-of-Distribution Failure through the Lens of Labeling Mechanisms: An Information Theoretic Approach
  • • On the data requirements of probing
  • • Neural reality of argument structure constructions
  • • Cool Books in 2021
  • • Quantifying the Task-Specific Information in Text-Based Classifications
  • • An unsupervised framework for tracing textual sources of moral change
  • • What do writing features tell us about AI papers?
  • • How is BERT surprised? Layerwise detection of linguistic anomalies
  • • NAACL 2021 papers
  • • Interpretable NLP 2021 winter
  • • Cool Books in 2020
  • • Examining the rhetorical capacities of neural language models
  • • An information theoretic view on selecting linguistic probes
  • • Trends in ACL 2020

Tag Cloud

ExplainableAI GAN Information NLP OOD PGM SVM Society VAE algorithms android bash book-review conferences database drone java js learning linguistics ml-basics os philosophy pytorch robotics teamwork web

Categories

  • Academia
  • Literature
  • Machine Learning
  • Philosophy
  • Software
© 2022 Zining Zhu
Powered by Hexo with Landscape.
Home Archives About