Browsing NeurIPS 2026 papers: three on evolutionary computation

Curious which papers are heading to NeurIPS 2026 this December and want an effortless way to browse the topics?

Jueun Kim 🙏 indexed the titles and abstracts of all accepted papers and built a remarkably clean, lightning-fast search tool so you can search and filter everything by topic or organisation.

I took a peek at the accepted papers and posters around Evolutionary Computation (a bit of an obsession of mine lately), and the search pulled up quite a fascinating batch.

Here are three that immediately caught my attention:

🔹 Evolutionary Feature Engineering for Structured Data: uses LLM-based evolution to automatically discover preprocessing transformations and compact feature programs for time-series and tabular data.
🔹 Automated Causal Effect Estimation through Self-Evolving AI: an evolutionary framework leveraging LLMs to discover and iteratively refine statistical causal inference methods.
🔹 Evolution Fine-Tuning: Learning to Discover Across 371 Optimization Tasks: teaches LLMs cross-task evolutionary search capabilities by fine-tuning on thousands of optimisation trajectories.

I’ll be covering them more extensively once I have time to read them more carefully!