Tag
This paper proposes Population-Curated Policy Optimization (PCPO), a sample-efficient parametric self-evolution method that helps LLM algorithm-design agents escape in-context evolutionary stagnation by training on self-generated algorithms. Trained on only 4 chip cases, PCPO outperforms OpenEvolve and ShinkaEvolve across 16 EDA cases, matches GPT-5.5 with an 8B model, and achieves 8.27× average speedup on GPU kernel design.
This paper establishes tight generalization bounds for multi-dimensional hyperparameter tuning in data-driven algorithm design, using real algebraic geometry and a multi-regime lower-bound framework to resolve theoretical gaps.
This paper introduces DyCA, a framework for LLM-assisted evolutionary search that uses dynamic instance clustering to improve tail robustness under heterogeneous instance distributions, outperforming existing LES baselines on four algorithm design tasks.
An introductory resource on parallel algorithms, covering fundamental concepts and techniques, from Carnegie Mellon University.
This paper introduces a framework for distribution-aware algorithm design where LLM agents learn to generate solver code specialized to target distributions, achieving high solution quality and significant speedups over standard solvers.
DeepMind announces AlphaEvolve, a Gemini-powered AI agent that combines large language models with automated evaluators to discover and optimize algorithms for mathematical and practical computing problems, improving efficiency in data centers, chip design, and AI training.