EMO-STA evolves a shared program archive across task families then adapts candidates to targets, outperforming matched-compute single-task evolution in most of eight families while reducing overfitting on low-data tasks like ARC.
Pawan and Dupont, Emilien and Ruiz, Francisco J
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
CONDITIONAL 2representative citing papers
Increasing LLM coding agents' reasoning effort raises cost and process complexity but does not reliably improve model quality across 140 controlled runs on networked anagram game data.
citing papers explorer
-
Evolutionary Multi-Task Optimization for LLM-Guided Program Discovery
EMO-STA evolves a shared program archive across task families then adapts candidates to targets, outperforming matched-compute single-task evolution in most of eight families while reducing overfitting on low-data tasks like ARC.
-
An Experimental Design Approach to Evaluating Agentic AI's Autonomous Model Discovery
Increasing LLM coding agents' reasoning effort raises cost and process complexity but does not reliably improve model quality across 140 controlled runs on networked anagram game data.