Papers arxiv:2605.09423 SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learning Published on May 10 Submitted by taesiri on May 12 Authors: , Xiaokang Ye , , , , , , Abstract SimWorld Studio presents an open-source platform using Unreal Engine 5 and … …
… A natural question arises: can frontier coding assistants reliably automate this evaluation process? Our study shows that simply prompting coding assistants is insufficient for this task. …
… The following papers were recommended by the Semantic Scholar API HATS: Hardness-Aware Trajectory Synthesis for GUI Agents 2026 WebCompass: Towards Multimodal Web Coding Evaluation for Code Language Models 2026 GameWorld: Towards Standardized and Verifiable Evaluation of Multimodal Game Agents 2026… …
… The following papers were recommended by the Semantic Scholar API AlphaEval: Evaluating Agents in Production 2026 Beyond Binary Correctness: Scaling Evaluation of Long-Horizon Agents on Subjective Enterprise Tasks 2026 Beyond Isolated Tasks: A Framework for Evaluating Coding Agents on Sequential So… …
… The following papers were recommended by the Semantic Scholar API Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses 2026 Synthesizing Multi-Agent Harnesses for Vulnerability Discovery 2026 AlphaLab: Autonomous Multi-Agent Research Across Optimization Do… …
Papers arxiv:2605.15040 Orchard: An Open-Source Agentic Modeling Framework Published on May 14 Submitted by Qianhui WU on May 15 Microsoft Research Authors: , , , , , Rui Yang , , , , , , , , Abstract Orchard is an open-source framework for scalable agentic modeling that enables training diverse au… …
…Each task averages roughly 8 minutes of wall-clock time and over 20 tool calls , and runs inside a reproducible Docker container hosting an actual CLI agent harness (OpenClaw, Claude Code, Codex…
… These developments lead to strong performance in multimodal coding, visual tool use , and framework-based agentic tasks, while preserving competitive text-only coding capability. …
… Experiments show that AgentRunbook-C achieves the best performance with 72.5% average accuracy, outperforming the strongest RAG baseline 48.5% and the off-the-shelf coding agent baseline 69.3% . Despite the strong performance gains, coding agent based methods have high latency costs. …
Papers arxiv:2605.14445 FrontierSmith: Synthesizing Open-Ended Coding Problems at Scale Published on May 14 Submitted by Qiuyang Mang on May 15 Authors: , , , , , , , , , , , , , , , , Abstract FrontierSmith automates the creation of open-ended coding problems from closed-ended tasks, improving LLM… …