• LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.

Changing a prompt or model can silently regress specific behaviors while the aggregate score holds. Grouping tests into behaviorally coherent slices and ranking them by regression risk tells you which behaviors broke under a fixed test budget.

• articleHow well LLM-based test generation techniques perform with newer LLM versions?Michael Konstantinou, Renzo Degiovanni, Mike Papadakis
• articleWhen Policies Change Probabilities: Modular Decision-Making for LLM Code ReviewRasvik Kudum, Max Corbett, Hitansh Paliwal, Romaisa Fatima, Thomas Jiralerspong, Sneheel Sarangi
• clipEvals encode failures you've already seenAI Engineer

Checking sign-in…

Loading comments…