01. Type-Safe Is Not Error-Free: A Constrained Decision Head Follows the Option Name, Not the Rubric Bound to It

88% worth reading · novelty 2.0/3 · benchmark/eval · cs.AI

02. Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents

87% worth reading · novelty 2.0/3 · new method · cs.AI

03. A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem

86% worth reading · novelty 1.9/3 · new method · cs.AI

04. Measuring the Serving Stack Instead of the Model: Hidden Confounds in Local Tool-Use Evaluation

86% worth reading · novelty 1.8/3 · benchmark/eval · cs.AI

05. MAGIC: Mixed-Granularity Agent Graphs via Incremental Construction with Dense-Reward Reinforcement Learning

83% worth reading · novelty 1.9/3 · new method · cs.LG

06. Governed AI-Agent Coordination for Dementia Care: Architecture, Safety Contracts, and Evidence-Derived Workflow Verification

83% worth reading · novelty 1.8/3 · new method · cs.MA

07. CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents

83% worth reading · novelty 1.8/3 · new method · cs.AI

08. When Does Execution Provenance Help Agent Memory Retrieval?

86% worth reading · novelty 1.6/3 · new method · cs.MA

09. Agensh: Scaling Organizational Intelligence to 1,024 Agents

82% worth reading · novelty 1.8/3 · new method · cs.CL

10. Knowledge Pull Requests for Continual Document Authoring

81% worth reading · novelty 1.9/3 · new method · cs.CL

11. Calibration Is Not Verification: Falsifiability-Aware Conformal Routing for Mixture-of-Agents

81% worth reading · novelty 1.8/3 · new method · cs.MA

12. SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving

80% worth reading · novelty 1.8/3 · benchmark/eval · cs.AI

13. Discovery-Driven Integration of Disjoint Tables via Text

78% worth reading · novelty 1.9/3 · new method · cs.CL

14. Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models

81% worth reading · novelty 1.7/3 · new method · cs.CL

15. Indirect tipping: a social attack surface in AI agent populations

76% worth reading · novelty 2.0/3 · new method · cs.MA

16. Qwen-Audio-Agent Technical Report

82% worth reading · novelty 1.3/3 · applied system · cs.MA

17. Behavior is Not Enough: A Mechanism-Based Evaluation of Social Norm Emergence in LLM Societies

74% worth reading · novelty 1.8/3 · benchmark/eval · cs.MA

18. A Task-Oriented Multi-Agent Framework for Complex Wearable Health Analysis

82% worth reading · novelty 1.1/3 · new method · cs.MA

19. SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue

74% worth reading · novelty 1.5/3 · new method · cs.AI

20. EquivSVA: A Formally Verified Dataset of Behavioral Assertions Across Equivalent RTL Implementations

67% worth reading · novelty 1.9/3 · benchmark/eval · cs.LG