arXiv:2608.27340v1 Announce Type: new Abstract: Steering interventions targeting eval-awareness, a model’s recognition that it is being tested, are…
Tag: cs.AI updates on arXiv.org
Mechanistic Reaction Prediction via Discrete Flow Matching on Graph-Structured Electron Occupation
arXiv:2608.27429v1 Announce Type: new Abstract: Chemical reactions are fundamentally transformations in electron space, yet most machine learning…
CorporateBench: Large-Scale Q&A Benchmarking with Temporal Knowledge Bases
arXiv:2608.27391v1 Announce Type: new Abstract: LLMs are increasingly able to answer complex questions about enterprise-scale document collections. But…
Learning a Continuous Sepsis Severity Score Without Hour-by-Hour Supervision: A Two-Site Retrospective Study
arXiv:2608.27421v1 Announce Type: new Abstract: Currently used sepsis severity indices rely on fixed variables and weights established decades ago, which…
What Makes Good Agentic Data? An ACE Lens on Data Generation for LLM Agents
arXiv:2608.27260v1 Announce Type: new Abstract: LLM agents increasingly rely on generated interaction data to learn how to interact with external…
LLMs Can Design Near-Optimal OR Algorithms
arXiv:2608.27296v1 Announce Type: new Abstract: We ask whether large language models (LLMs) can design effective algorithms for well-specified operations…
Verify Smarter, Evolve Further: Efficient Harness Evolution through Behavior-Aware Verification
arXiv:2608.27311v1 Announce Type: new Abstract: Agent harnesses shape how language-model agents use instructions, tools, and runtime components, but…
BrailleBench: Investigating Multi-Criteria Braille Comprehension in Large Language Models
arXiv:2608.27268v1 Announce Type: new Abstract: Although Large language models (LLMs) mediate access to knowledge and computational assistance, their…
Naive Prompt Optimization: Rethinking the Need for Complex Prompt Search
arXiv:2608.27266v1 Announce Type: new Abstract: Efficiently improving autonomous agents across diverse tasks is central to accelerating recursive…
BPMN4CAI: A BPMN Extension for Modeling Dynamic Conversational AI
arXiv:2608.27149v1 Announce Type: new Abstract: Conversational AI systems, such as chatbots and virtual assistants, are becoming increasingly important to…