Search:
Match:
2 results
Paper#llm🔬 ResearchAnalyzed: Jan 3, 2026 06:15

Classifying Long Legal Documents with Chunking and Temporal

Published:Dec 31, 2025 17:48
1 min read
ArXiv

Analysis

This paper addresses the practical challenges of classifying long legal documents using Transformer-based models. The core contribution is a method that uses short, randomly selected chunks of text to overcome computational limitations and improve efficiency. The deployment pipeline using Temporal is also a key aspect, highlighting the importance of robust and reliable processing for real-world applications. The reported F-score and processing time provide valuable benchmarks.
Reference

The best model had a weighted F-score of 0.898, while the pipeline running on CPU had a processing median time of 498 seconds per 100 files.

Research#XAI🔬 ResearchAnalyzed: Jan 10, 2026 13:07

Explainable AI Powers Smart Greenhouse Management: A Deep Dive into Interpretability

Published:Dec 4, 2025 19:41
1 min read
ArXiv

Analysis

This research explores the application of explainable AI (XAI) in the context of smart greenhouse control, focusing on the interpretability of a Temporal Fusion Transformer. Understanding the 'why' behind AI decisions is critical for adoption and trust, particularly in agricultural applications where environmental control is paramount.
Reference

The research investigates the interpretability of a Temporal Fusion Transformer in smart greenhouse control.