Never miss a new edition of The Variable, our weekly newsletter featuring a top-notch selection of editors’ picks, deep dives, community news, and more.
Fine-tuning? RAG? Chain-of-thought? We suspect that for many of our readers, these LLM-optimization approaches—as relevant as they might still be—feel a tad stale.
If you’d like to catch up on cutting-edge topics in the sprawling world of large language models, read on. This week’s Variable highlights three recent articles that will help you create powerful LLM workflows and overcome emerging challenges.
How to Create an LLM Judge That Aligns with Human Labels
Evaluating the quality of LLM outputs continues to be a thorn in many a practitioner’s side. Elena Samuylova presents a lucid, hands-on guide to building a robust LLM-as-a-judge pipeline that produces reliable and consistent results.
Your 1M+ Context Window LLM Is Less Powerful Than You Think
Before you worry about how many tokens your model can process, consider its effective working memory. Tobias Schnabel explains why.
Exploring Prompt Learning: Using English Feedback to Optimize LLM Systems
Based on her team’s recent work, Aparna Dhinakaran outlines a promising new approach that “uses natural language feedback to iteratively improve prompts.”
This Week’s Most-Read Stories
Catch up on the articles our community has been buzzing about in recent days:
Topic Model Labelling with LLMs, by Petr Koráb
Accuracy Is Dead: Calibration, Discrimination, and Other Metrics You Actually Need, by Pol Marin
The Future of AI Agent Communication with ACP, by Mariya Mansurova
Other Recommended Reads
From anomaly detection to self-evolving AI, our authors continue to cover fascinating topics in data science and machine learning. Here are a few more must-reads to keep you busy:
- I Analysed 25,000 Hotel Names and Found Four Surprising Truths, by Anna Gordun Peiro
- Don’t Waste Your Labeled Anomalies: 3 Practical Strategies to Boost Anomaly Detection Performance, by Shuai Guo
- The Age of Self-Evolving AI Is Here, by Moulik Gupta
- Midyear 2025 AI Reflection, by Marina Tosic
- Evaluation-Driven Development for LLM-Powered Products: Lessons from Building in Healthcare, by Robert Martin-Short
Meet Our New Authors
Explore top-notch work from some of our recently added contributors:
- Shireesh Kumar Singh is an IBM Cloud software engineer whose first TDS articles focus on network-congestion forecasting and knowledge graphs.
- Pavel Timonin joins us with software-engineering expertise of his own; his debut story is a hands-on computer vision deep dive.
We love publishing articles from new authors, so if you’ve recently written an interesting project walkthrough, tutorial, or theoretical reflection on any of our core topics, why not share it with us?
