Prompt EngineeringMert Yuksekgonul et al. (Stanford)Stanford University

TextGrad published

TextGrad published — backpropagates natural-language feedback as gradients to optimize prompts, code, molecules and treatment plans

Why it matters

Extended the automatic-differentiation metaphor to text systems in general

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2406.07496

Added in Week 35, 2026 · See that week

Same week

SB Intuitions releases Sarashina1 series 7B/13B/65B Japanese LLMsOfficialMaxime Labonne publishes AbliterationOfficialOpenVLAPaperReconciling Kaplan and Chinchilla Scaling LawsPaper

Nearby on the same axis

The Prompt Report publishedPaperMIPRO publishedPaperLet's Think Dot by Dot publishedPaperIEEE Spectrum publishes AI Prompt Engineering Is DeadPress