Large Language Model (LLM) agents in distributed decision-making scenarios often operate as spatially separated entities and coordinate from fragmented local information. However, this physical dispersion constrains agents to limited local perception and finite temporal horizons. We characterize this bottleneck as spatiotemporal partial observability, in which spatial and temporal limitations are fundamentally coupled. Resolving spatial conflicts requires temporal reasoning about neighbors’ future actions, while temporal planning requires spatial context beyond local perception. To bridge this gap, we introduce **MACRO-LLM**, LLM-empowered multi-agent collaborative reasoning under spatiotemporal partial observability. The architecture integrates spatial and temporal reasoning within each decision cycle via three interdependent modules: (1) the CoProposer mitigates *temporal uncertainty* by screening candidate actions via LLM-predicted rollouts; (2) the Negotiator overcomes *spatial myopia* by resolving conflicts through mean-field statistical aggregation, grounded in the CoProposer’s LLM-estimated rollout rewards; and (3) the Introspector closes the reasoning loop by analyzing environmental drift and attributing performance changes to refine strategies. Extensive evaluations on two complex long-horizon tasks, cooperative platoon planning and pandemic control, demonstrate that our framework enables robust coordination under spatiotemporal partial observability.

LLM agents rely on long-term interaction histories to tailor their responses to individual users. However, historical personalization signals may become stale or misaligned with expectations implied by the current context, leading agents to follow outdated preferences during reasoning. We refer to this failure mode as personalization inertia, which shifts the agent’s behavior away from the user’s current intent. Existing methods mainly optimize final responses or memory retrieval, leaving such failures underexplored. In this paper, we propose Personalization Inertia VerificatiOn and RegulaTion (Pivot), a framework for mitigating personalization inertia in long-term interactions with agents. Concretely, we construct a trajectory-independent evidence condition as a diagnostic reference, model prefix-evidence consistency to assess compatibility with the current personalization evidence, and learn a rollback policy to determine whether and where to intervene. The selected intervention triggers local repair, preserving the reliable prefix while regenerating only the potentially inertia-affected suffix under evidence-guided constraints. Experiments on benchmark datasets demonstrate consistent gains in personalization accuracy and trajectory faithfulness with targeted intervention overhead.

LLM-empowered paper-code discrepancy detection has received growing concern since the scaling of research submissions exceeds the manual review capability. However, the limited context capacity and one-sided discrepancy detection of existing single-agent LLM paradigms lead to an inferior recall performance in detecting discrepancies. In this paper, we propose **Dude**, the first **Du**al-**De**tection Multi-Agent System for paper-code discrepancy detection. We discover that the granularity asymmetry of the paper-language and code-language introduces over-interpretation and over-reporting challenges in a multi-agent system design for discrepancy detection, resulting in increasing false positives. To address this, we propose a granularity-aligned negotiation and a two-stage salience-filtering mechanism in Dude, which effectively prevents agents from falsely reporting discrepancies. Experimental results in real-world paper-code discrepancy datasets showcase Dude’s significant recall and precision improvement by up to **22.8%**, increasing F1 score by up to **18.7%** compared to baseline methods.

Autonomous LLM agents increasingly operate in long-horizon, interactive settings where success depends on reusing experience accumulated over extended histories. However, existing agent memory systems are fundamentally constrained by text-context budgets: storing or revisiting raw trajectories is prohibitively token-expensive, while summarization and text-only retrieval trade token savings for information loss and fragmented evidence. To address this limitation, we propose Optical Context Retrieval Memory (OCR-Memory), a memory framework that leverages the visual modality as a high-density representation of agent experience, enabling retention of arbitrarily long histories with minimal prompt overhead at retrieval time. Specifically, OCR-Memory renders historical trajectories into images annotated with unique visual identifiers. OCR-Memory retrieves stored experience via a locate-and-transcribe paradigm that selects relevant regions through visual anchors and retrieves the corresponding verbatim text, avoiding free-form generation and reducing hallucination. Experiments on long-horizon agent benchmarks show consistent gains under strict context limits, demonstrating that optical encoding increases effective memory capacity while preserving faithful evidence recovery.

In large language model (LLM) agents, reasoning trajectories are treated as reliable internal beliefs for guiding actions and updating memory. However, coherent reasoning can still violate logical or evidential constraints, allowing unsupported beliefs repeatedly stored and propagated across decision steps, leading to systematic behavioral drift in long-horizon agentic systems. Most existing strategies rely on the consensus mechanism, conflating agreement with faithfulness. In this paper, inspired by the vulnerability of unfaithful intermediate reasoning trajectories, we propose Self-Audited Verified Reasoning (SAVeR), a novel framework that enforces verification over internal belief states within the agent before action commitment, achieving faithful reasoning. Concretely, we structurally generate persona-based diverse candidate beliefs for selection under a faithfulness-relevant structure space. To achieve reasoning faithfulness, we perform adversarial auditing to localize violations and repair through constraint-guided minimal interventions under verifiable acceptance criteria. Extensive experiments on six benchmark datasets demonstrate that our approach consistently improves reasoning faithfulness while preserving competitive end-task performance.