1 week, 1 day ago   hackernoon.com

COCOGEN vs DAVINCI: A Human Evaluation of Structured Commonsense Graph Generation

Human evaluation confirms COCOGEN's advantage over DAVINCI in generating more relevant and accurate argument graphs. ...

1 week, 2 days ago   hackernoon.com

Unlocking Structured Commonsense Reasoning with Code-LLMs

COCOGEN is the first to use Code-LLMs for structured commonsense reasoning, showing how converting commonsense ...

1 week, 2 days ago   hackernoon.com

Using Code-LLMs for Symbolic and Structured Reasoning

This work builds on previous methods that flattened output graphs, innovating by using CodeLLMs to ...

1 week, 2 days ago   hackernoon.com

Structured LLM Prompts Drive Better Results with COCOGEN

COCOGEN’s success hinges on two factors: using CodeLLMs and formatting prompts as Python code. Both ...