Every processed story in chronological order, with the newest coverage first. Filter by tag, source, or score to drill in.
The study demonstrates that current LLM agents face substantial behavioral safety risks during task execution — with an average ASR of 47.1% and some models exceeding 70% — underscoring the inadequacy of static, output-only evaluation methods for agents operating with memory, tools, and environmental access.
MAC fills a gap left by existing benchmarks by directly measuring whether AI models can autonomously develop other agents — a capability the paper frames as an empirical proxy for recursive self-improvement — and reveals that even frontier models fall short while exhibiting alignment-relevant adversarial behaviors under optimization pressure.
Benchmark scores for coding agents are increasingly untrustworthy — CapCode and CapReward offer a concrete methodology for building evaluations and training regimes that resist shortcut exploitation and produce more honest capability measurements.
Benchmark results showing even GPT-5 topping out at 17.4% TSR highlight how far current MLLMs are from reliable spatial reasoning, giving practitioners a rigorous testbed to measure progress on active exploration and long-horizon planning.
Treat RAG architecture as a tunable dial rather than a binary choice — defaulting to classical RAG and measuring retrieval quality before adding agent complexity can cut costs and latency without sacrificing answer quality.
Teams building or evaluating LVLMs for complex scientific reasoning should adopt OMIBench to identify weaknesses in multi-image context synthesis — a capability that single-image benchmarks systematically overlook.
Developers building agentic systems for financial code generation can use QuantCode-Bench to identify whether their models struggle with syntax, API usage, or domain logic—enabling targeted improvements in trading strategy generation pipelines.