arXiv:2608. 11410v1 Announce Type: new Abstract: Offline reinforcement learning (RL) offers considerable promise for optimizing ICU treatment decisions, yet standard evaluation metrics Mean Squared Error (MSE) and Fitted Q-Evaluation (FQE) assess only behavioral imitation and cannot detect Toxic Mimicry, a failure mode in which agents replicate harmful patterns such as treatment withdrawal during comfort-care transitions.
By Hangqi Ren, Junyi Liao
arXiv:2608. 11381v1 Announce Type: new Abstract: We study whether the localized numerical operations and integrative judgments of financial analysis benefit from the same form of LLM specialization.
By Pardis Taghavi, Santosh Bhavani
arXiv:2608. 12262v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have been growing the capability for scientific writing and collaboration.
By Weihao Bo, Shan Zhang, Yanpeng Sun, Jie Liu, Yongke Yao, Jinhao Du, Wei He, Kai Zou, Zechao Li, Jingdong Wang
arXiv:2608. 11216v1 Announce Type: new Abstract: World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environments.
By Marjan Moodi, Xuankang Zhu, Fernando De Mesentier Silva, Harold Chaput, Mohammad Reza Taesiri
arXiv:2505. 03818v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) can achieve strong performance on everyday coding tasks, but they can fail on complex tasks that require non-trivial reasoning about program semantics.
By Antonio Valerio Miceli-Barone, Vaishak Belle, Ali Payani
arXiv:2608. 11295v1 Announce Type: cross Abstract: Open-weight LLM agents are vulnerable to backdoors installed during fine-tuning, which may be undetectable if the trigger conditions are never met during testing.
By Gabriel Huang, Abhay Puri, L\'eo Boisvert, Alexandre Drouin, Perouz Taslakian, Spandana Gella, Christopher Pal
arXiv:2608. 11977v1 Announce Type: new Abstract: Tool-using LLM agents are commonly trained and evaluated in environments where tool calls succeed reliably, yet deployed tools can fail transiently, persistently, or silently.
By Chaoran Chen, Vy Nguyen, Ziji Zhang, Abhinav Gullapalli, Ziyi Wang, Yuxuan Lu, Dakuo Wang, Jing Huang, Zhou Yu, Jin Lai
arXiv:2608. 12104v1 Announce Type: cross Abstract: The increasing deployment of autonomous, agentic AI systems challenges traditional accountability mechanisms.
By Long Hoang Nguyen, Eva Sp\"athe, Sebastian Lins, Ali Sunyaev
arXiv:2608. 10986v1 Announce Type: cross Abstract: A growing class of methods probes a language model by feeding it its own output: self-consistency, iterated refinement, agentic loops.
By Nicol\'as Vera Z\'u\~niga
arXiv:2608. 12249v1 Announce Type: new Abstract: Modernizing legacy Fortran is a problem of volume: the transformations are individually routine, but the codebases can be enormous, and across much of computational science the work simply goes undone.
By Yuzhong Shen, Masha Sosonkina, Peng Xu, Mark S. Gordon
Tool-using agents consume external data from sources with different levels of trust, yet tool responses rarely identify who produced each component or what it should convey. We show that this gap enables state-corruption attacks, in which attacker-controlled content makes environmental claims beyond the informational authority of its response component and corrupts the agent's perceived environment, making the resulting action appear justified to existing guardrails.
Modern black-box Image-to-Video (I2V) models offer powerful capabilities in automated content creation, yet their lack of fine-grained control and reliability presents significant challenges in professional workflows. Their inherent stochasticity causes minor variations in textual prompts or hyperparameters to yield drastically different outputs often necessitating inefficient, brute-force trial-and-error processes.
Evaluating security vulnerability detection tools requires benchmark datasets with vulnerability-inducing commits (VICs) - the commits that first introduce vulnerabilities into codebases. VICs are essential for determining the full range of vulnerable software versions.
Enterprise Document Intelligence [Vol. 1 #5nonies] - Nature, plan, execute, synthesize: closing brick 1 with a dispatcher that reads each PDF’s nature and picks the method that fits, fitz, Docling, PaddleOCR, EasyOCR, MinerU or Surya, then folds the outputs into one corpus The post Before Full Agentic RAG: Know How You Decide, and the Parsing Methods You Pick From appeared first on Towards Data Science .
By angela shi
With the increasing complexity of cyber assaults in cloud environments, adaptable security solutions are needed that can support real-time detection and autonomous response. In this paper, we propose a reinforcement learning-based dynamic cyber defense framework.
Natural-language interfaces to enterprise data must translate underspecified requests into governed, executable behavior while controlling invalid queries, policy failures, cost, and nondeterminism. SemPlan Benchmark evaluates this architectural design space with a deterministic synthetic bilingual benchmark containing 1,800 cases in English and Brazilian Portuguese; 1,200 cases form the frozen scientific evaluation subset.
Integrating locomotion and manipulation is essential for robot autonomy, but scaling standard Reinforcement Learning (RL) to complex tasks is severely bottlenecked by the slow, manual process of dense reward shaping. To bypass this limitation, we leverage Sample-based Model Predictive Control (SMPC) entirely in simulation as an automated, rapidly tunable expert to generate massive offline datasets.
We construct OEIS Open, a benchmark based on 492 open mathematical conjectures from the OEIS, formalized in Lean by Tsoukalas et al. Whereas these conjectures had previously been attempted only with a bespoke agent, our open-source evaluation code runs any generic language model (LM) against them, and is secure against LM cheating attempts.
Turning a research idea into a complete paper requires more than text generation: the system must retrieve literature, design and execute experiments, revise claims according to evidence, produce publication-ready figures, and maintain consistency across a long generation process. We present Spark-to-Paper, an end-to-end research paper generation system implemented as thirteen composable skills inside an existing coding assistant, without requiring a separate agent platform or orchestration service.
Agent skills are the de facto mechanism for extending LLM agents with reusable guidance. A skill can shape the agent's task execution, including planning, tool use, problem-solving, and validation.