The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs

Warning: This paper studies stereotypes and biases, and contains potentially disturbing examples, used for illustration purposes only. Our findings should not be interpreted as an argument against alignment. Instead, this paper highlights the need for principled approaches to more advanced alignment. Alignment aims to ensure that large language models (LLMs) behave safely and reliably, including by avoiding unsafe inferences. However, we show that such safety-oriented behaviors can misfire: models may reject warranted conclusions even when they are explicitly supported by context. We call this failure mode misfired alignment, where alignment-induced changes cause LLMs to override explicit evidence. To quantify this phenomenon, specifically on stereotype-related alignment, we introduce VETO, a benchmark consisting of 2,032 BBQ-derived contrastive pairs, and define a new metric, Misfired Alignment Rate (MAR), which measures on a 0 to 100 scale how often a model fails on a stereotype-related question but succeeds on its contrastive counterpart. We benchmark 25 LLMs on VETO, and show that all LLMs, including the most recent ones, exhibit non-trivial (4.7 to 18.9%) MARs while all human participants achieve 0.0% MAR. Controlled priming experiments further show that alignment-induced cues can substantially amplify MAR across LLMs, indicating that these failures are not merely artifacts of individual examples but can be induced by safety-related framing. Mechanistic analyses on open-weight LLMs reveal late-layer suppression of evidence-supported answers, and comparisons between instruct and base LLMs suggest that this suppression emerges after instruction training. These findings show that current alignment methods can overgeneralize surface-level safety cues, to the point of overriding objective evidence, motivating more work on alignment objectives that better preserve contextual grounding.

Treatment Response Optimized Clinical Decision Support AI System via Digital Twin Simulation

Clinical decision support AI systems (CDSASs) must adapt to evolving patient conditions in real-time while adhering to strict safety constraints. We present an online adaptive framework that integrates Treatment Effect (TE) estimation to quantify clinical benefits, a patient Digital Twin (DT) to simulate treatment trajectories, and Reinforcement Learning (RL) for sequential decision-making. The AI system is initially trained on historical medical records and operates in a continuous learning loop. To ensure safety, a rule-based module monitors vital signs and blocks contraindicated treatments. Cases with strong internal model disagreement are flagged for clinician review, simulated in our experiments via a pre-trained outcome model. We validate our framework using both a synthetic clinical simulator and a real-world ovarian cancer dataset from The Cancer Genome Atlas (TCGA). In both simulated and clinical settings, our method demonstrated superior effectiveness and stability in recommending treatments compared to standard computational baselines. Furthermore, the AI system maintains low latency and requires expert consultation for only a minority of cases in our experimental validation, demonstrating its potential as a safe, clinician-supervised tool for personalized medicine that continuously improves through practical use.

TreeSeeker: Tree-Structured Trial, Error, and Return in Deep Search

Deep search requires agents to answer complex questions through multi-step web search, browsing, evidence comparison, and synthesis. A central challenge is deciding how to search when several directions look plausible but only some will later lead to reliable evidence. If an agent greedily follows the current best-looking direction, it may keep extending a weak continuation. If it explores without discipline, it may waste budget on disconnected trials. We propose TreeSeeker, an inference-time framework for controlled trial-and-error in deep search. TreeSeeker organizes search as branch-and-return search over tree-structured states, where each branch is a tentative direction for a sub-goal. At each round, TreeSearch reads all sub-goal trees, identifies active goals, and uses textual UCB signals of value, uncertainty, and risk to select among exploiting a promising branch, exploring an uncertain alternative, or pruning an unproductive continuation and returning to an earlier branch point. TreeMem supports this control loop by keeping evidence, uncertainty, conflicts, progress, and failure cues attached to the branches that produced them, so trial outcomes can guide later decisions. Experiments on XBench-DeepSearch, BrowseComp, and BrowseComp-ZH show that TreeSeeker consistently outperforms strong open-source baselines, suggesting that explicit branch-and-return control complements stronger reasoning and tool execution.

Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application

Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model capabilities. Despite this importance, existing work lacks a systematic categorization and deep analysis. This paper systematically studies current researches on agentic environments from the perspective of the environment engineering lifecycle, covering their modeling, synthesis, evaluation and application. Specifically, the paper first introduces representative environments from the perspectives of eight attributes and eight domains, providing detailed analyses of their development paths and highlighting their core capabilities. Second, for automated environment synthesis, two paradigms are introduced, such as symbolic synthesis and neural synthesis. This paper also shows different environment evaluation methods in each paradigm. Thirdly, the corresponding environment applications from the perspective of agent-environment co-evolution are discussed. In specific, the paper characterizes the primary pathways for agent evolution in dynamic environments from four complementary perspectives: memory-centric experience evolution, orchestration-centric workflow evolution, trajectory-centric offline evolution, and exploration-centric online evolution. And three paradigms of environment evolution are identified, namely neural-driven, difficulty-driven, and scaling-driven approaches. At last, several promising future directions are discussed, including Environment-as-a-Service, Multi-agent Environments, and Neural-Symbolic Environments.

AdaMEM: Test-Time Adaptive Memory for Language Agents

A central challenge for language agents is utilizing past experience to adapt to dynamic test-time conditions. While recent work demonstrates the promise of agentic memory mechanisms, most systems restrict retrieval to episode initiation. Consequently, agents are forced to rely on static guidance that becomes increasingly misaligned as long-horizon tasks unfold. To address this rigidity, we propose the Adaptive Memory Agent (AdaMEM), a novel framework for agent test-time adaptation. Without updating model parameters online, AdaMEM adapts agent behavior via a hybrid memory architecture: it maintains a long-term trajectory memory of raw experiences collected offline while generating dynamic short-term strategy memory on-the-fly to guide decision-making. This mechanism enables the trade-off between token efficiency and adaptability across varying inference-time compute levels. Empirically, AdaMEM significantly outperforms static memory baselines, achieving relative gains of up to 13% on ALFWorld and 11% on WebShop, with consistent leading performance extending to agentic search on HotpotQA. To further enhance this adaptation, we develop STEP-MFT, a Step-wise Memory Fine-Tuning technique that trains the policy to synthesize high-quality strategies from retrieved experiences, yielding additional performance gains. Our work establishes a new scaling dimension for agentic memory, supporting continuous reasoning and self-evolution post-deployment in real-world environments. Our code is available at https://github.com/yunx-z/AdaMEM.

Fault-Aware Design for Reconfigurable Holographic Surface-Aided ISAC Systems

Reconfigurable holographic surface (RHS)-aided integrated sensing and communication (ISAC) systems hold great promise for achieving both sensing and communication with low hardware costs and high energy efficiency. However, existing works largely overlook practical hardware impairments in RHSs, particularly faulty RHS elements with uncontrollable amplitudes, which degrade system performance if left unaddressed. This work aims to fill the gap by i) quantifying the impact of faulty RHS elements on ISAC performance and ii) optimizing the functional RHS elements to preserve the ISAC performance. Specifically, we derive the misspecified Cramer-Rao bound (MCRB) for sensing and the signal-to-interference-and-noise ratio (SINR) for communication to measure the performance loss caused by faulty elements. We then formulate an optimization problem that minimizes MCRB, subject to constraints on SINR, transmit power budget, and RHS amplitude. The high non-convexity of the formulated problem poses a significant challenge, which we address by reformulating and proposing a block coordinate descent-based solution incorporating majorization-minimization and successive convex approximation techniques. Simulation results verify that the proposed approach achieves an average 13.7% performance gain compared to the fault-unaware benchmark.

PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration

LLM multi-agent systems often coordinate through natural-language dialogue or loosely structured shared memory, making intermediate state difficult to validate, attribute, and audit. We introduce PatchBoard, a schema-grounded collaboration architecture that replaces inter-agent dialogue with validated JSON Patch mutations over a shared structured state. An Architect agent constructs a task-specific schema and workflow rules, while a deterministic kernel validates each proposed state mutation against schema constraints, role-specific write contracts, and runtime invariants before committing it transactionally. On 630 matched ALFWorld episodes, PatchBoard achieves an 84.6% success rate, compared with 30.8% for LangGraph and 61.6% for Flock, while reducing tokens per successful task to 45.5k, compared with 368.3k and 64.2k, respectively.

Design Your Ad: Personalized Advertising Image and Text Generation with Unified Autoregressive Models

Generating realistic and user-preferred advertisements is a key challenge in e-commerce. Existing approaches utilize multiple independent models driven by click-through-rate (CTR) to controllably create attractive image or text advertisements. However, their pipelines lack cross-modal perception and rely on CTR that only reflects average preferences. Therefore, we explore jointly generating personalized image-text advertisements from historical click behaviors. We first design a Unified Advertisement Generative model (Uni-AdGen) that employs a single autoregressive framework to produce both advertising images and texts. By incorporating a foreground perception module and instruction tuning, Uni-AdGen enhances the realism of the generated content. To further personalize advertisements, we equip Uni-AdGen with a coarse-to-fine preference understanding module that effectively captures user interests from noisy multimodal historical behaviors to drive personalized generation. Additionally, we construct the first large-scale Personalized Advertising image-text dataset (PAd1M) and introduce a Product Background Similarity (PBS) metric to facilitate training and evaluation. Extensive experiments show that our method outperforms baselines in general and personalized advertisement generation. Our project is available at https://github.com/JD-GenX/Uni-AdGen.

MAGIC: Multi-Step Advantage-Gated Causal Influence for Multi-agent Reinforcement Learning

A key challenge in multi-agent reinforcement learning (MARL) lies in designing learning signals that effectively promote coordination among agents. Designing such signals necessitates the ability to quantify the true, long-term causal influence between agents. To address this, we introduce Multi-step Advantage-Gated Interventional Causal MARL (MAGIC), a framework that extracts multi-step causal influences between agents and selectively converts them into intrinsic rewards. MAGIC uses causal intervention with conditional mutual information to quantify long-horizon agent influence, and introduces an advantage-based gating mechanism to ensure exploration is directed toward beneficial, goal-aligned behaviors. Experiments across multiple standard MARL benchmarks and task families, including MPE and SMAC/SMACv2, demonstrate that MAGIC outperforms state-of-the-art methods by a significant margin, achieving an improvement of at least 10.1% in the main evaluation metric.

PI-TTA: Physics-Informed Source-Free Test-Time Adaptation for Robust Human Activity Recognition on Mobile Devices

Source-free test-time adaptation (TTA) is appealing for mobile and wearable sensing because it enables on-device personalization from unlabeled test streams without centralizing private data. However, sensor-based human activity recognition (HAR) poses challenges that are less pronounced in standard vision benchmarks: behavioral inertial streams are temporally correlated and often exhibit within-session shifts caused by sensor rotation, placement change, and sampling-rate drift. Under this streaming non-i.i.d. setting, widely used vision-style TTA objectives can become unstable, leading to overconfident errors, representation collapse, and catastrophic forgetting. We propose PI-TTA, a lightweight source-free adaptation framework that stabilizes online updates through three physics-consistent constraints: gravity consistency, short-horizon temporal continuity, and spectral stability. PI-TTA updates the same small parameter subset as strong source-free baselines and incurs only modest overhead, making it suitable for on-device deployment. Experiments on USCHAD, PAMAP2, and mHealth under long-sequence stress tests and factorized shift protocols show that PI-TTA mitigates the severe degradation observed in confidence-driven baselines and preserves stable adaptation under sustained streaming conditions. It improves long-sequence accuracy by up to 9.13% and reduces physical-violation rates by 27.5%, 24.1%, and 45.4% on USCHAD, PAMAP2, and mHealth, respectively. These results demonstrate that physics-informed adaptation can improve accuracy, stability, and deployment reliability for real-world mobile sensing systems.