SeisEvo LLM-Driven Agent Search for Seismic Algorithms
August 20, 2026
SeisEvo uses an LLM-driven multi-agent search to evolve standalone, white-box seismic reconstruction algorithms. Instead of training a neural network, the system modifies components of classical algorithms while enforcing physical constraints and scoring candidates by execution.
HOW THIS AFFECTS YOU
●
researcherYou can automate the search for interpretable, physics-constrained algorithms using LLM agents.