Negated Complementary Commonsense using Large Language Models

1 code implementation • 13 Jul 2023 • Navid Rezaei, Marek Z. Reformat

Larger language models, such as GPT-3, have shown to be excellent in many tasks.

Reinforcement Learning for Topic Models

1 code implementation • 8 May 2023 • Jeremy Costello, Marek Z. Reformat

We apply reinforcement learning techniques to topic modeling by replacing the variational autoencoder in ProdLDA with a continuous action space reinforcement learning policy.

Reinforcement Learning (RL) Sentence Embeddings +1

Utilizing Language Models to Expand Vision-Based Commonsense Knowledge Graphs

1 code implementation • journal 2022 • Navid Rezaei, Marek Z. Reformat

It means the triples are verbalized to natural language, and after being processed, the results are converted back to triples and added to the commonsense knowledge graph.

Knowledge Graphs Language Modelling

Super-Prompting: Utilizing Model-Independent Contextual Data to Reduce Data Annotation Required in Visual Commonsense Tasks

no code implementations • 25 Apr 2022 • Navid Rezaei, Marek Z. Reformat

To evaluate our results, we use a dataset focusing on visual commonsense reasoning in time.

Few-Shot Learning In-Context Learning +2

Analysis of Word Embeddings Using Fuzzy Clustering

no code implementations • 17 Jul 2019 • Shahin Atakishiyev, Marek Z. Reformat

In data dominated systems and applications, a concept of representing words in a numerical format has gained a lot of attention.

Clustering Semantic Similarity +2