Competitive programmers collaborated with Sakana AI’s ShinkaEvolve to win 1st place in the 2025 ICFP Programming Contest 👏 sakana.ai/icfp-2025 What happens when competitive programming experts master a cutting-edge AI tool? Team Unagi, which includes Sakana AI's Research Scientist Takuya Akiba (@iwiwi), recently deployed “ShinkaEvolve”—our evolutionary code optimization framework released last month—in the famous global programming contest: @icfpcontest2025. Amazingly, the result was a 1st place finish. 🥇 This human-AI collaboration also yielded several fascinating insights. ShinkaEvolve automatically improved the team’s SAT encoding, accelerating computation speed by up to 10x. This enabled them to find near-optimal solutions for large-scale problems that were previously intractable. Moreover, it created a virtuous cycle where the AI’s discovery of using auxiliary variables became a significant hint for subsequent human-led development. This case shows how a complementary collaboration between humans and AI can be a powerful approach for the future: a cycle where humans design the high-level strategy, AI performs the intensive improvements, and then humans interpret the AI’s discoveries to apply them to new challenges. A huge congratulations to everyone on Team Unagi! At Sakana AI, we look forward to fostering more collaborations between human professionals and cutting-edge AI.

Competitive programmers collaborated with Sakana AI’s ShinkaEvolve to win 1st place in the 2025 ICFP Programming Contest 👏 sakana.ai/icfp-2025 What happens when competitive programming experts master a cutting-edge AI tool? Team Unagi, which includes Sakana AI's Research Scientist Takuya Akiba (@iwiwi), recently deployed “ShinkaEvolve”—our evolutionary code optimization framework released last month—in the famous global programming contest: @icfpcontest2025. Amazingly, the result was a 1st place finish. 🥇 This human-AI collaboration also yielded several fascinating insights. ShinkaEvolve automatically improved the team’s SAT encoding, accelerating computation speed by up to 10x. This enabled them to find near-optimal solutions for large-scale problems that were previously intractable. Moreover, it created a virtuous cycle where the AI’s discovery of using auxiliary variables became a significant hint for subsequent human-led development. This case shows how a complementary collaboration between humans and AI can be a powerful approach for the future: a cycle where humans design the high-level strategy, AI performs the intensive improvements, and then humans interpret the AI’s discoveries to apply them to new challenges. A huge congratulations to everyone on Team Unagi! At Sakana AI, we look forward to fostering more collaborations between human professionals and cutting-edge AI.

We’re excited to introduce ShinkaEvolve: An open-source framework that evolves programs for scientific discovery with unprecedented sample-efficiency. Blog: sakana.ai/shinka-evolve/ Code: github.com/SakanaAI/ShinkaEv… Like AlphaEvolve and its variants, our framework leverages LLMs to find state-of-the-art solutions to complex problems, but using orders of magnitude fewer resources! Many evolutionary AI systems are powerful but act like brute-force engines, burning thousands of samples to find good solutions. This makes discovery slow and expensive. We took inspiration from the efficiency of nature. ‘Shinka’ (進化) is Japanese for evolution, and we designed our system to be just as resourceful. On the classic circle packing optimization problem, ShinkaEvolve discovered a new state-of-the-art solution using only 150 samples. This is a big leap in efficiency compared to previous methods that required thousands of evaluations. We applied ShinkaEvolve to a diverse set of hard problems with real-world applications: 1/ AIME Math Reasoning: It evolved sophisticated agentic scaffolds that significantly outperform strong baselines, discovering an entire Pareto frontier of solutions trading performance for efficiency. 2/ Competitive Programming: On ALE-Bench (a benchmark for NP-Hard optimization problems), ShinkaEvolve took the best existing agent's solutions and improved them, turning a 5th place solution on one task into a 2nd place leaderboard rank in a competitive programming competition. 3/ LLM Training: We even turned ShinkaEvolve inward to improve LLMs themselves. It tackled the open challenge of designing load balancing losses for Mixture-of-Experts (MoE) models. It discovered a novel loss function that leads to better expert specialization and consistently improves model performance and perplexity. ShinkaEvolve achieves its remarkable sample-efficiency through three key innovations that work together: (1) an adaptive parent sampling strategy to balance exploration and exploitation, (2) novelty-based rejection filtering to avoid redundant work, and (3) a bandit-based LLM ensemble that dynamically picks the best model for the job. By making ShinkaEvolve open-source and highly sample-efficient, our goal is to democratize access to advanced, open-ended discovery tools. Our vision for ShinkaEvolve is to be an easy-to-use companion tool to help scientists and engineers with their daily work. We believe that building more efficient, nature-inspired systems is key to unlocking the future of AI-driven scientific research. We are excited to see what the community builds with it! Learn more in our technical report: arxiv.org/abs/2509.19349

We are excited to share that “Continuous Thought Machines” has been accepted as a Spotlight at #NeurIPS2025! 🧠✨ The CTM is an AI that mimics biological brains by using neural dynamics & synchronization to think over time. It can solve complex mazes by building internal maps, gaze around images to classify them, and learn algorithms—all emergent from its core design. This is just the beginning. A hint of what we're exploring next… (video attached!) The team: @LearningLukeD @ciaran_regan_ @risi1979 @jeffreyseely @YesThisIsLion

We are excited to share that “Continuous Thought Machines” has been accepted as a Spotlight at #NeurIPS2025! 🧠✨ The CTM is an AI that mimics biological brains by using neural dynamics & synchronization to think over time. It can solve complex mazes by building internal maps, gaze around images to classify them, and learn algorithms—all emergent from its core design. This is just the beginning. A hint of what we're exploring next… (video attached!) The team: @LearningLukeD @ciaran_regan_ @risi1979 @jeffreyseely @YesThisIsLion