cover

How CODEX Model Size Influences COCOGEN's Output Quality

24 Apr 2025

Exploring the impact of model size on COCOGEN's performance, CODEX-002 outperforms CODEX-001. Prompt sensitivity decreases as model size increases.

cover

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

24 Apr 2025

Human evaluation shows COCOGEN outperforms DAVINCI in generating more relevant and correct commonsense graphs for tasks like EXPLAGRAPHS and PROSCRIPT.

cover

Using Code-LLMs for Structured Commonsense Reasoning

24 Apr 2025

COCOGEN pioneers the use of Code-LLMs for structured commonsense generation, opening new avenues for NLP tasks that require structured prediction and reasoning.

cover

Unlocking Structured Commonsense Reasoning with Code-LLMs

23 Apr 2025

COCOGEN pioneers the use of Code-LLMs for structured commonsense generation, opening new avenues for NLP tasks that require structured prediction and reasoning.

cover

Using Code-LLMs for Symbolic and Structured Reasoning

23 Apr 2025

COCOGEN uses CodeLLMs for structured commonsense generation, going beyond traditional symbolic reasoning tasks by translating them into Python code.

cover

Structured LLM Prompts Drive Better Results with COCOGEN

23 Apr 2025

COCOGEN’s performance gains stem from combining structured code prompts with CodeLLMs—outperforming text-based models even under dynamic or duplicate inputs.

cover

COCOGEN Sets Few-Shot Benchmark in Entity and Argument Graph Tasks

22 Apr 2025

COCOGEN delivers top-tier results in entity state tracking and argument graph generation—surpassing fine-tuned models with just a few Python-coded examples.

cover

Study Shows Few-Shot Code Generation Outperforms Fine-Tuned Models

22 Apr 2025

COCOGEN beats GPT-3 and fine-tuned models in structured commonsense tasks with just 15 Python-based examples. Efficient, powerful, and precise.

cover

Why Converting Graphs to Python Code Improves AI Reasoning

22 Apr 2025

COCOGEN converts commonsense graphs into Python code, helping CodeLLMs outperform traditional models in structured reasoning tasks.