… View arXiv page View PDF Add to collection Community AI agents will increasingly need to interact with unfamiliar AI counterparts in settings like negotiation and bargaining. In this paper, we ask: can we predict the next decision of a black-box AI agent after seeing only a few prior interactions? …
… Compared to previous benchmarks on agentic web browsing , AutoResearchBench is distinguished along three dimensions: it is research-oriented, calling for in-depth comprehension of scientific concepts; literature-focused, demanding fine-grained utilization of detailed information; and open-ended, in… …
… The following papers were recommended by the Semantic Scholar API DataClaw: A Process-Oriented Agent Benchmark for Exploratory Real-World Data Analysis 2026 World Reasoning Arena 2026 FutureWorld: A Live Reinforcement Learning Environment for Predictive Agents with Real-World Outcome Rewards 2026 P… …
Papers arxiv:2605.01214 Agentic AI Systems Should Be Designed as Marginal Token Allocators Published on May 2 Submitted by siqi zhu on May 5 University of Illinois at Urbana-Champaign Authors: Siqi Zhu Abstract Agentic AI systems should be evaluated as marginal token allocation economies rather tha… …
… View arXiv page View PDF GitHub 15 Add to collection Community that multi-layer evaluation stack in AcademiClaw is the most interesting bit for me, because it turns a single pass/fail into a landscape of diagnostics from code execution to browser end-to-end checks. i'd love to see an ablation where… …
Papers arxiv:2604.20441 MedSkillAudit: A Domain-Specific Audit Framework for Medical Research Agent Skills Published on Apr 22 Submitted by Huimei Wang on May 7 AIPOCH Authors: , , Huimei Wang , , , , , , , , , , Abstract A domain-specific audit framework for medical research agent skills demonstra… …
… No offline indexing. 📊The Results: DCI outperforms top baselines across 13 benchmarks, with average gains of: 🔍 Agentic Search BrowseComp-Plus : +11.0% 🧠 Multi-hop QA: +30.7% 📈 IR Ranking: +21.5% 💡 Insights: Beyond accuracy, we conduct a series of controlled ablation studies to pinpoint the sources… …
… AI-generated summary Workspace learning requires AI agents to identify, reason over, exploit, and update explicit and implicit dependencies among heterogeneous files in a worker's workspace, enabling them to complete both routine and advanced tasks effectively. …
… The following papers were recommended by the Semantic Scholar API AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents 2026 FACT: Compositional Kernel Synthesis with a Three-Stage Agentic Workflow 2026 Kernel-Smith: A Unified Recipe for Evolutionary Kernel Optimizat… …
… We present a comprehensive empirical study of multi-turn agentic RL for medical AI, built on , a gymnasium-compatible environment spanning 10 clinical domains with 3.6K+ tasks, 135 domain-specific tools, and a knowledge base of 828K medical passages. …