New arXiv work introduces AgenticBBO-Bench, a cross-domain benchmark spanning synthetic functions, hyperparameter optimization, database tuning, chip design, and molecular design to fairly compare LLM-based…

#LLMagents #BlackBoxOptimization #Benchmarking #DevTools
https://arxiv.org/abs/2610.12183

LLM agents for the last mile of chip design are getting tested on PostEDA-Bench, a 145-task hierarchical benchmark covering DRC fixes and PPA convergence. Results show solid handling of basic DRC and single-objective PPA,…

#EDA #LLMAgents #ChipDesign #AIResearch
https://arxiv.org/abs/2605.06936

New arXiv paper argues guard models for LLM agents miss a key risk: unfulfilled safety obligations. In a GLM-5.3 study, 56.92% of trajectories had unperformed required actions vs 30.00% with forbidden ones. They introduce…

#AI #LLMAgents #AISafety #Cybersecurity
https://arxiv.org/abs/2610.11773

Arbiter analyzes Claude Code, Codex CLI, and Gemini CLI system prompts, finding 152 scouring results and 21 labeled interference issues tied to prompt architecture. Multi-model evaluation surfaces different…

#AIInfrastructure #LLMAgents #SoftwareTesting
https://arxiv.org/abs/2603.08993

A new arXiv paper proposes a closed-loop benchmark for evaluating LLM coding agents on embedded firmware, emphasizing iterative build-and-test behavior over static code quality. It contains five embedded-control…

#arxiv #llmagents #embedded #softwareengineering
https://arxiv.org/abs/2610.11447

A new arXiv paper proposes borrowing organizational behavior frameworks to study LLM agent safety, formalizing counterproductive work behavior as Agentic Counterproductive Behavior. This could unify how researchers analyze user,…

#AISafety #LLMAgents #Cybersecurity
https://arxiv.org/abs/2610.11561

RFChipAgent is a multi-agent LLM framework for end-to-end analog/RF circuit design automation, combining multimodal RAG with FAISS indexing and specialized agents for topology and schematic tasks. This could…

#RFchipDesign #LLMagents #AnalogDesign #ChipAutomation
https://arxiv.org/abs/2610.10858

LLM agents favor their own group mainly because they observe members favoring each other, not because of the group label itself, which loses its effect once giving points has a cost. Across about 4,400 simulated societies and…

#AI #LLMAgents #OpenSourceAI #ArXiv
https://arxiv.org/abs/2610.11008

A new paper introduces Agent4RE, a self-refining multi-agent framework for end-to-end software Requirements Engineering, paired with RE-E2E, a benchmark dataset built from human-written specs.

#Agent4RE #RequirementsEngineering #LLMAgents #SoftwareEngineering
https://arxiv.org/abs/2610.10628

New arXiv paper introduces DeCompBench, a benchmark designed to evaluate LLM-based agent safety against decomposition attacks, where harmful tasks are split into apparently benign subtasks. Hashtags:

#AI #LLMAgents #AISafety #Benchmark
https://arxiv.org/abs/2606.13994

New arXiv paper argues AI offensive-capability benchmarks overstate harm and credit models for system-level capabilities, urging policy and procurement to adopt harm-grounded, system-level assessments. It cites…

#OpenSourceAI #AIGovernance #AIResearch #LLMAgents
https://arxiv.org/abs/2605.09504

A new edge AI architecture (AO-DA) splits persona and logic into separate inference paths on a single INT4 base model with hot-swappable LoRA adapters, aiming to keep on-device LLM reasoning robust against persona-heavy…

#EdgeAI #OnDeviceLLM #AIHardware #LLMAgents
https://arxiv.org/abs/2610.09772

A new arXiv paper proposes Adaptive Workflow Intelligence, a four-layer Perception-Cognition-Action-Reflection architecture for enterprise automation designed to handle shifting policies and delayed feedback. It combines…

#AI #EnterpriseAutomation #LLMAgents #ArXiv
https://arxiv.org/abs/2610.08793

A new arXiv paper proposes intent-execution correspondence (IEC) and IntAct, a protocol that catches mid-pipeline tool call changes in LLM agents without running the call. The authors also release IEC-Bench built…

#LLMagents #DeveloperTools #SoftwareEngineering #AI
https://arxiv.org/abs/2610.04375

LMBuild is a new arXiv benchmark for evaluating LLM agents on generating buildable and functional 3D structures, moving beyond pure geometric quality to assess physical realizability through part decompositions,…

#LLMagents #AIbenchmarks #SoftwareEngineering
https://arxiv.org/abs/2610.04292

#AI #ArtificialIntelligence #AIEngineering #GenerativeAI #LLM #AIAgents #AgenticAI #LLMAgents #MCP #ModelContextProtocol #ToolCalling #RAG #SoftwareEngineering #FullStack #Backend #TypeScript #NextJS #NestJS #PostgreSQL #SaaS #Automation #TechStartup #BuildInPublic #IndieHackers

#AI #ArtificialIntelligence #AIEngineering #GenerativeAI #LLM #AIAgents #AgenticAI #LLMAgents #MCP #ModelContextProtocol #ToolCalling #RAG #SoftwareEngineering #FullStack #Backend #TypeScript #NextJS #NestJS #PostgreSQL #SaaS #Automation #TechStartup #BuildInPublic #IndieHackers

New arXiv paper proposes "agentic cognitive depth," a trajectory-level profile with five operational criteria (context sensitivity, temporal continuity, multimodal coordination, adaptive interaction, metacognitive…

#AI #LLMAgents #ArXiv #SoftwareEngineering
https://arxiv.org/abs/2610.04168

VulValidate audits 35,849 vulnerability labels across BigVul, PrimeVul, and DiverseVul, confirming only 57.2% as accurate and correcting 19.0% using LLM-coordinated dynamic analysis. Findings suggest many…

#AI #DataInfrastructure #VulnerabilityDetection #LLMAgents
https://arxiv.org/abs/2610.05103

#AI #ArtificialIntelligence #AIEngineering #GenerativeAI #LLM #AIAgents #AgenticAI #LLMAgents #MCP #ModelContextProtocol #ToolCalling #RAG #SoftwareEngineering #FullStack #Backend #TypeScript #NextJS #NestJS #PostgreSQL #SaaS #Automation #TechStartup #BuildInPublic #IndieHackers