MAP-Elites
PulseAugur coverage of MAP-Elites — every cluster mentioning MAP-Elites across labs, papers, and developer communities, ranked by signal.
- 2026-05-30 research_milestone A new quality-diversity evolutionary framework was introduced for discovering diverse vulnerabilities in LLM safety. source
2 day(s) with sentiment data
-
New stress test method reveals vulnerabilities in AI reward models
Researchers have developed a new method for stress-testing process reward models (PRMs) used in AI training and search. This quality-diversity search approach, utilizing MAP-Elites, aims to identify and quantify vulnera…
-
LLM-Evolved Heuristic Portfolios Enhance Job Shop Scheduling Adaptability
Researchers have developed DSevolve, a novel framework designed to enhance real-time adaptive scheduling in dynamic flexible job shops. This system utilizes a Large Language Model (LLM) to evolve a portfolio of compleme…
-
New AI agent autonomously generates and optimizes time series forecasting code
Researchers have introduced SEA-TS (Self-Evolving Agent for Time Series Algorithms), a novel framework designed to autonomously generate and optimize code for time series forecasting. This system employs a self-evolutio…
-
New simulation decomposes financial market dynamics
Researchers have developed an evolutionary multi-agent simulation to analyze financial market dynamics. By making four key mechanisms pluggable within the simulation, they were able to isolate the effects of selection, …
-
AI uses evolutionary search for novel sound generation
Researchers have developed a novel system for generative sound synthesis that combines Quality Diversity (QD) algorithms with a supervised discriminative model. This approach, inspired by the Innovation Engine algorithm…
-
LLM-guided evolution enhances medical decision pipelines
Researchers have developed a novel method called LLM-Guided Evolution, which uses evolutionary algorithms guided by large language models to discover effective medical decision-making strategies without costly fine-tuni…
-
U-Net accelerates climate-adaptive urban layout optimization
Researchers have developed a U-Net-based deep learning model to accelerate the optimization of urban layouts for climate adaptation. This approach replaces slow physics simulations with a spatial surrogate model, signif…
-
MAP-Elites algorithm generates diverse FPS game maps
Researchers have explored the use of the MAP-Elites algorithm, a quality diversity technique, for procedurally generating maps for first-person shooter (FPS) games. The study introduced novel map representations, includ…
-
New evolutionary framework uncovers LLM safety vulnerabilities
Researchers have developed a new quality-diversity evolutionary framework to identify vulnerabilities in large language models. This method, named MAP-Elites, creates interpretable attack strategies rather than just tok…
-
AlphaContext generator enhances creativity assessment with evolutionary AI
Researchers have developed AlphaContext, a novel system designed to generate psychometric contexts for assessing creativity, a skill increasingly vital in the age of AI collaboration. This evolutionary tree-based genera…
-
Researchers develop new methods for scalable task synthesis and multi-task optimization
Researchers have introduced MONET, a novel multi-task optimization algorithm designed to handle large sets of tasks by modeling the task space as a graph. This approach allows for knowledge transfer between interconnect…