…A Multi-Agent LLM Framework with Specialized Analyzers (2026) AnomalyAgent: Training-Free Agentic Models for Zero-/Few-Shot Anomaly Detection (2026) IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools (2026…
…Hyeon Hwang , , Abstract ThinC framework enables mathematical problem solving where code serves as the primary reasoning mechanism instead of a verification tool, demonstrating superior performance on math benchmarks. AI-generated summary Tool…
…Controlled Benchmark Rewriting and Analogical Reasoning for Deceptive Out-of-Distribution Scenarios (2026) Who Tests the Testers? Systematic Enumeration and Coverage Audit of LLM Agent Tool Call Safety (2026) Visual Confused Deputy…
…Repurposing Git Commit Messages as a Structured Knowledge Protocol for AI Coding Agents (2026) Test-Driven AI Agent Definition (TDAD): Compiling Tool-Using Agents from Behavioral Specifications (2026) Project Prometheus: Bridging the…
…In this paper, we equip these context optimizers with Wikipedia search and browser tools for active information seeking . We show that naively adding these tools to a standard sequential context optimization pipeline…
…Rather than relying on an LLM to directly produce a final scene, Function2Scene performs iterative evaluation and refinement through a tool-augmented check-and-repair loop , combining geometric measurements , LLM-based contextual…
…Extensive experiments demonstrate that our method significantly outperforms existing baselines on fine-grained benchmarks, while maintaining the flexibility to support tool-assisted inference workflows. View arXiv page View PDF Project page GitHub…
…Likewise, since video technologies became widely adopted, video has served not only as a powerful tool for recording and communicating visual information, but also as a vehicle for emotions, attitudes, and social…
…raises CWE aggregation from 72.9\% to 81.1\% at 16K; and on GPQA with agentic tool use, it yields a 24\% relative gain in pass@1. Attention analysis shows that FocuSFT…
…diversity, difficulty, and quality, using model internals extracted with Sparse Autoencoder (SAE), an advanced mechanistic interpretability tool. Each property grounds a concrete data engineering operation: SAE-space clustering with moderate batch mixing…