virtual-engineers-revolutionize-machine-learning-with-optimized-recommendations
draws-on arxiv-2606-03841v1
draws-on arxiv-2606-03221v1
draws-on arxiv-2606-03143v1
headline Virtual Engineers Revolutionize Machine Learning with Optimized Recommendations." Researchers introduce EvoDS, a self-evolving autonomous data science agent that learns to expand its skills and adaptively manage long-term context through agentic reinforcement learning.
finds EvoDS outperforms state-of-the-art open-source data science agents by an average of 28.9% across four diverse benchmarks.
finds The agent's hierarchical design reduces tool-selection error and its optimization objective aligns with an information bottleneck principle.
finds EvoDS's code and data are available on GitHub.
finds VirtualMLE, an LLM-agent framework, leverages the cognitive capabilities of Large Language Models to organize recommender optimizing into a closed loop of execution, reflection, and memory update.
finds VirtualMLE reaches competitive recommendation quality with substantially fewer trials on three Amazon SR benchmarks.
finds Cognition summaries distilled from previous datasets can significantly accelerate the search process on unseen datasets.
finds VirtualMLE's codes are available.
finds FederatedSkill, a privacy-preserving framework for collaborative agent evolution, utilizes semantic skill diffs to dynamically model client-specific capability boundaries.
finds FederatedSkill achieves substantial gains over self-evolving baselines, with up to a 44.4% increase in success rate and a 37.5% reduction in computational cost.
finds The framework is evaluated across 20 distinct agent task families.
Also as Turtle and JSON-LD.
← all subjects