论文检索

Applications · Robotics

Kunjal Panchal, Saayan Mitra, Sunav Choudhary, Victor Bursztyn, Somdeb Sarkhel, Hui Guan

LLM-based multi-agent embodied planning remains impractical due to prohibitively high execution latency. We identify failed actions as the dominant bottleneck, stemming from two core challenges: inaccurate state tracking under partial observability and inefficient coordination that produces redundant or conflicting actions. We introduce Mosaic, a runtime-efficient multi-agent planning framework that addresses both challenges. Mosaic maintains accurate yet lightweight state tracking through agent-centric semantic memory that stores objects in relative coordinates, enabling geometric transformations and coordination. It ensures efficient coordination through Integer Linear Programming that allocates actions at every planning step, enforcing physical feasibility and inter-agent coordination constraints. Across AI2-THOR and search-and-rescue benchmarks, Mosaic achieves 27–32% faster execution, 30–33% fewer LLM calls, 25–31% fewer steps, and 4–10% points higher success rates. These results demonstrate that efficient memory and constraint-guided coordination are critical for scalable, low-latency multi-agent planning.

Deep Learning · Large Language Models

Boyuan Xiao, Bohong Chen, Yumeng Li, Ji Feng, Yao-Xiang Ding, Kun Zhou

In embodied vision-language decision making tasks such as robotic manipulation and navigation, Vision-Language and Vision-Language-Action Models (VLMs \& VLAs) are powerful tools with different benefits: VLMs are better at long-term planning, while VLAs are better at reactive control. However, their performance is limited by the same perceptual bottleneck: visual hallucinations arise due to the models’ inability to distinguish task-relevant objects from distractors. In principle, accurate identification and focus on critical objects while filtering out irrelevant ones is the key to break this limitation. A straightforward solution is one-step focus: directly attending to essential objects. However, this approach proves ineffective because effective focus inherently requires deep scene understanding. To this end, we propose ${\it SceneDiver}$, a coarse-to-fine focus plan generation method for VLMs leveraging their long-term planning abilities, that first constructs a holistic scene graph to establish initial comprehension, then progressively decomposes the task into simpler sub-problems through an iterative cycle of recognition, understanding, and analysis. To enable reactive control, we also design a lightweight adapter for distilling the deliberate focus ability into VLAs. Evaluations on standard embodied AI benchmarks confirm that our method substantially reduces visual hallucinations for both VLMs and VLAs, while preserving computational efficiency in tasks requiring fast execution.

Probabilistic Methods · Everything Else

Amirmohammad Farzaneh, Salvatore D'oro, Osvaldo Simeone

Large language model (LLM)-powered agents can translate high-level user intents into plans and actions in an environment. Yet after observing an outcome, users may wonder: What if I had phrased my intent differently? We introduce a framework that enables such counterfactual reasoning in agentic LLM-driven control scenarios, while providing formal reliability guarantees. Our approach models the closed-loop interaction between a user, an LLM-based agent, and an environment as a structural causal model (SCM), and leverages test-time scaling to generate multiple candidate counterfactual outcomes via probabilistic abduction. Through an offline calibration phase, the proposed conformal counterfactual generation (CCG) yields sets of counterfactual outcomes that are guaranteed to contain the true counterfactual outcome with high probability. We showcase the performance of CCG on a wireless network control use case, demonstrating significant advantages compared to naive re-execution baselines.

Deep Learning · Large Language Models

Yue Pei, Hongming Zhang, Jiarui Guan, Jusheng Zhang, Liang Lin, Haogang Zhu, Ziliang Chen

Large language models (LLMs) are increasingly used as planners for cooperative embodied agents, but multi-agent settings amplify inconsistency under partial observability and make explicit communication costly or even unavailable. Many existing approaches rely on online message passing; when communication is removed, agents often fall back to independent local planning that suffers from tacit miscoordination. We introduce Contextual Tacit Communication, a training-free protocol that aligns decentralized decisions with a joint LLM value score without explicit message actions. Our method measures context-conditioned value rectifications via residual banding to pinpoint miscoordination actions and amortizes the resulting coordination signals into a retrieval-augmented Tacit Rule Memory that provides prompt-level cooperation rules at execution time. Experiments on VIKI, C-WAH, and TDW-MAT show that our approach improves cooperation performance over baselines while reducing runtime overhead compared with communication-based methods.

Deep Learning · Large Language Models

Jing Huang, Lidong Zhang, Mutian Bao, Yadong Li, Xingzhong Xu, Jinjian Zhang, Jie Liu, Ming Kong, Qiang Zhu

The Automated Design of Multi-Agent Systems (Auto-MAS) has emerged as a promising framework for addressing complex reasoning tasks. However, existing approaches often suffer from structural rigidity and entangle the design of system topology with the implementation of individual agents. To overcome these limitations, we propose MAS-Architect, a framework that automates MAS design through a novel code-based declarative MAS paradigm rooted in the \textit{Separation of Concerns} principle. By decoupling topology planning from node implementation via a unified interface, our approach enables the from-scratch generation of task-adaptive architectures. We further employ a \textit{Distill-then-Explore} training strategy to optimize these designs. Comprehensive experiments on five benchmarks show that MAS-Architect sets a new Pareto frontier in the efficiency–performance trade-off: it surpasses state-of-the-art methods while substantially lowering token usage. Notably, the framework achieves a strong average accuracy of 78.7\% across benchmarks with an inference cost of only 2,533 tokens per query. Qualitative analysis reveals the autonomous emergence of advanced collaboration patterns, validating the generative flexibility of the declarative paradigm. Code and data will be released.

Deep Learning · Sequential Models, Time series

Hao Miao, Ni Zhang, Zefei Ning, Li Wang

Battery lifetime early prediction is crucial for safety assessment and decision planning. However, early-stage degradation signals are extremely weak and difficult to distinguish from noise. Existing methods primarily rely on denoising or signal decomposition, which risks losing critical degradation cues. In nature, most organisms exhibit binaural effect, exploiting differences between left and right auditory inputs to enhance perceptual reliability. Inspired by this, we propose DITING, a weak degradation listener for battery lifetime early prediction. DITING first employs optimal-transport-based selective matching to extract a robust health template from initial cycles for degradation representation. To manifest degradation signals from noise, we further design a tri-coupled degradation manifestation mechanism. By exploiting the randomness of noise, matched responses under symmetric coupling suppress stochastic fluctuations. Conversely, cumulative deviations driven by degradation propagate through the coupling to form stable bilateral discrepancies. This design effectively amplifies weak cues in the early stage for lifetime prediction. Experiments on multiple datasets demonstrate that DITING achieves state-of-the-art performance and provides more reliable early support for full-lifecycle battery management. The code is available at https://anonymous.4open.science/r/Batt_DITING.

Social Aspects · Security

Shanghao Shi, Xiao Wang, Chaoyu Zhang, Hao Li, Wenjing Lou, Thomas Hou, Yevgeniy Vorobeychik, Chongjie Zhang, Ning Zhang

The integration of external tools has substantially expanded the capabilities of large language model (LLM) agents, but also introduced new attack surfaces beyond prompt injection. In particular, cross-tool description poisoning can manipulate planner-visible tool metadata to steer an agent’s trajectory, even if the poisoned tool itself is never chosen. To understand the effectiveness of existing attacks against this emerging threat, we evaluate several existing agent defenses against prompt-injection and find they transfer poorly to cross-tool description poisoning. Building on this insight, we propose Tool-Guard, a novel defense based on a new concept called isolated planning, in which tool invocations that are detected as misaligned or suspicious cause the corresponding tool to be placed in a quarantined list (the influenced list), breaking further influence from poisoned descriptions. With this influence isolated, the tool can continue to be used to support the task, enabling a robust defense that preserves legitimate tool utility. Experiments on the AgentDojo and ASB benchmarks show that \sysname substantially reduces attack success while maintaining high task utility.

Theory · Reinforcement Learning and Planning

Wanqiao Xu, Allen Nie, Ruijie Zheng, Aditya Modi, Adith Swaminathan, Ching-An Cheng

Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language model (LLM) agents. While impressive empirical demonstrations have been shown, so far a principled framing of these decision problems remains lacking. In this paper, we formalize the Learning from Language Feedback (LLF) problem, assert sufficient assumptions to enable learning despite latent rewards, and introduce *transfer eluder dimension* as a measure to characterize the hardness of LLF problems. We formalize the intuition that information in the feedback governs the learning complexity of LLF problems. We demonstrate cases where learning from rich language feedback can be exponentially faster than learning from reward. We develop a no-regret algorithm, called `HELiX`, that provably solves LLF problems through sequential interactions, with performance guarantees that scale with the transfer eluder dimension of the problem. Across several empirical domains, we show that `HELiX` performs well even when repeatedly prompting LLMs does not work reliably. Our contributions mark an important step towards designing principled interactive learning algorithms from generic language feedback.

Zhen Wang, Fan Bai, Zhongyan Luo, Jinyan Su, Kaiser Sun, Xinle Yu, Jieyuan Liu, Kun Zhou, Claire Cardie, Mark Dredze 等

Autonomous agents powered by large language models (LLMs) promise to accelerate scientific discovery, but rigorously evaluating their capacity for verifiable discovery remains a central challenge. Existing benchmarks face a trade-off: they either rely on LLM-as-judge evaluations of automatically generated papers, or optimize isolated performance metrics that provide only coarse proxies for scientific insight. To address this, we introduce FIRE-Bench (Full-cycle Insight Rediscovery Evaluation), a benchmark that evaluates agents through the rediscovery of established findings from recent, high-impact machine learning research. Agents are given only a high-level research question from a published study and must autonomously design experiments, implement code, execute their plans, and derive conclusions supported by empirical evidence. We evaluate a range of state-of-the-art agents with frontier model backbones, such as gpt-5, on FIRE-Bench. Our results show that full-cycle scientific research remains challenging for current agent systems: even the strongest agents achieve limited rediscovery success, exhibit high variance across runs, and display recurring failure modes in experimental design, execution, and evidence-based reasoning. Overall, FIRE-Bench provides a rigorous and diagnostic framework for measuring progress toward reliable agent-driven scientific discovery.

Reinforcement Learning · Planning

Michael Psenka, Michael Rabbat, Aditi Krishnapriyan, Yann LeCun, Amir Bar

World models simulate environment dynamics from raw sensory inputs like video. However, using them for planning can be challenging due to the vast and unstructured search space. We propose a robust and highly parallelizable planner that leverages the differentiability of the learned world model for efficient optimization, solving long-horizon control tasks from visual input. Our method treats states as optimization variables ("virtual states") with soft dynamics constraints, enabling parallel computation and easier optimization. To facilitate exploration and avoid local optima, we introduce stochasticity into the states. To mitigate sensitive gradients through high-dimensional vision-based world models, we modify the gradient structure to descend towards valid plans while only requiring action-input gradients. Our approach can be viewed as a stochastic version of a non-condensed or collocation-based optimal controller. We provide theoretical justification and experiments on video-based world models, where our resulting planner outperforms existing planning algorithms like the cross-entropy method (CEM) and vanilla gradient-based optimization (GD) on long-horizon experiments, both in success rate and time to convergence.

Reinforcement Learning · Planning

Michael Katz, Harsha Kokel, Christian Muise, Shirin Sohrabi, Sarath Sreedharan

In over sixty years since its inception, the field of planning has made significant contributions to both the theory and practice of building planning software that can solve a never-before-seen planning problem. This was done through established practices of rigorous design and evaluation of planning systems. **It is our position that this rigor should be applied to the current trend of work on planning with large language models.** One way to do so is by correctly incorporating the insights, tools, and data from the automated planning community into the design and evaluation of LLM-based planners. The experience and expertise of the planning community could play a crucial role in accelerating the development of LLM-based planners. This position is particularly important in light of the abundance of recent works that replicate and propagate the same pitfalls that the planning community has encountered and learned from. We believe that establishing practices that avoid such known pitfalls will contribute greatly to the progress in building LLM-based planners and to planning in general.

Reinforcement Learning · Batch/Offline

Anthony Kobanda, Waris Radji, Odalric-Ambrym Maillard, Rémy Portelas

Goal-Conditioned Reinforcement Learning aims to design agents that can reach specified goals, notably from previously collected trajectories in the offline setting. In this context, graph-based approaches have been proposed to mitigate compounding value-estimation errors in long-horizon navigation tasks. However, existing approaches typically rely on dense keypoint coverage of the dataset support, resulting in computationally expensive planning. Moreover, they lack explicit mechanisms to adapt to topological changes (e.g., new obstacles), hindering deployment in live applications such as video game environments. To address these two shortcomings, we introduce Adaptive Quasimetric Mapping (AQM), an offline framework leveraging a “time-to-reach” quasimetric learned from the available data. Crucially, it builds a sparse cover of the dataset support, as a greedy approximation to a dominating set problem. At test-time, the resulting graph is carefully pruned by comparing the observed edge traversal time against a time-to-reach budget derived from the quasimetric, thus enabling zero-shot replanning. Empirically, we evaluate AQM on navigation tasks ranging from a classical to a video-game-like benchmark evaluating adaptation across tasks. We show that AQM achieves competitive performance while requiring up to 100× fewer keypoints than prior approaches, hence demonstrating the relevance of topological abstraction for goal-conditioned navigation.

Applications · Robotics

Xiaoliang Fan, jiarui chen, Zhuodong Liu, Ziqi Yang, Peixuan Xu, Ruimin Shen, Junhui Liu, Jianzhong Qi, Cheng Wang

Embodied AI (EAI) systems are rapidly transitioning from simulations into real-world domestic and other sensitive environments. However, recent EAI solutions have largely demonstrated advancements within \emph{isolated stages} such as instruction, perception, planning and interaction, without considering their coupled privacy implications in high-frequency deployments where privacy leakage is often irreversible. This position paper argues that optimizing these components independently creates a systemic privacy crisis when deployed in sensitive settings, thereby advancing the position that privacy in EAI is a life cycle-level architectural constraint rather than a stage-local feature. To address these challenges, we propose Secure Privacy Integration in Next-generation Embodied AI~(\textbf{SPINE}), a unified privacy-aware framework that treats privacy as a dynamic control signal governing \emph{cross-stage} coupling throughout the entire EAI life cycle. SPINE decomposes the EAI pipeline into various stages and establishes a multi-criterion privacy classification matrix to orchestrate contextual sensitivity across stage boundaries. We conduct preliminary simulation and real-world case studies to conceptually validate how privacy constraints propagate downstream to reshape system behavior, illustrating the insufficiency of fragmented privacy patches and motivating future research directions into secure yet functional embodied AI systems.

Fan Feng, Yujia Zheng, Minghao Fu, Yongqiang Chen, Guangyi Chen, Kevin Murphy, Biwei Huang, Kun Zhang

Learning and planning in imagination using world models provides an effective paradigm for training agents for decision-making. However, existing approaches often rely on high-dimensional latent spaces or generic visual embeddings that retain many factors irrelevant to control, limiting efficiency and generalization across tasks. To this end, we study how agents can learn world models with representations that are task-specific, minimal, and sufficient for decision making. We achieve this via a closed-loop synergy between the agent and the world model, in which structured world-model learning distills task-sufficient representations from informative interaction data. On the agent side, agents perform active probing of the environment to collect informative trajectories that expose task-relevant latent factors, guided by an adaptive curriculum. On the world-model side, we learn structured representations over observations to distill compact, task-sufficient latent states from the collected interaction data. This synergy enables the recovery of task-sufficient latent representations that capture all control-relevant factors empirically. Leveraging these representations, the resulting policies achieve improved sample efficiency and systematic generalization, including generalization across skills, object–skill compositions, and previously unseen tasks on standard continuous control and robotic manipulation benchmarks.

Deep Learning · Large Language Models

Lin Ai, Victor Bursztyn, Xiang Chen, Julia Hirschberg, Saayan Mitra

Recent advances in large language models (LLMs) have enabled deep research systems that synthesize comprehensive, report-style answers to open-ended queries by combining retrieval, reasoning, and generation. Yet, most frameworks rely on rigid workflows with one-shot scoping and long autonomous runs, offering little room for course correction if user intent shifts mid-process. We present **SteER**, a framework for steerable deep research that introduces interpretable, mid-process control into long-horizon research workflows. At each decision point, **SteER** uses a cost–benefit formulation to determine whether to pause for user input or proceed autonomously. It combines diversity-aware planning with utility signals that reward alignment, novelty, and coverage, and maintains a live persona model that evolves throughout the session. **SteER** outperforms state-of-the-art open-source and proprietary baselines by up to 22.80% on alignment, leads on quality metrics such as breadth and balance, and is preferred by human readers in 85%+ of pairwise alignment judgments. We also introduce a persona–query benchmark and data-generation pipeline. To our knowledge, this is the first work to advance deep research with an interactive, interpretable control paradigm, paving the way for controllable, user-aligned agents in long-form tasks.

Deep Learning · Generative Models and Autoencoders

Kaisen Yang, Jayden Teoh, Kaicheng Yang, Yitong Zhang, Alex Lamb

Masked Diffusion Models (MDMs) offer greater flexibility in decoding order than autoregressive models but require careful path planning to achieve high-quality generation. While existing samplers typically greedily select positions with the lowest uncertainty, we identify their fundamental limitations through failure case analysis, showing they overlook the impact of current actions on subsequent steps and fail to optimize cumulative uncertainty. To bridge this gap, we propose the **Info-Gain Sampler**, a principled decoding framework that balances immediate costs with information gain. Our method features a simple but effective objective and an efficient implementation that ensures practical overhead is minimal. Extensive evaluations across diverse architectures and tasks (reasoning, coding, creative writing, and image generation) demonstrate that Info-Gain Sampler consistently outperforms existing samplers, significantly raising the performance ceiling of MDMs. For instance, it achieves a 5.5\% improvement in average accuracy on reasoning tasks and a 63.1\% win-rate in creative writing; notably, on reasoning tasks it reduces cumulative uncertainty from 78.4 to 48.6, outperforming the best baseline by a large margin.

Theory · Reinforcement Learning and Planning

Chenye Yang, Weiyu Xu, Lifeng Lai

Offline Reinforcement Learning from Human Feedback (RLHF) pipelines such as Direct Preference Optimization (DPO) train on a pre-collected preference dataset, which makes them vulnerable to preference poisoning attack. We study label flip attacks against log-linear DPO. We first illustrate that flipping one preference label induces a parameter-independent shift in the DPO gradient. Using this key property, we can then convert the targeted poisoning problem into a structured binary sparse approximation problem. To solve this problem, we develop two attack methods: Binary-Aware Lattice Attack (BAL-A) and Binary Matching Pursuit Attack (BMP-A). BAL-A embeds the binary flip selection problem into a binary-aware lattice and applies Lenstra-Lenstra-Lovász reduction and Babai's nearest plane algorithm; we provide sufficient conditions that enforce binary coefficients and recover the minimum-flip objective. BMP-A adapts binary matching pursuit to our non-normalized gradient dictionary and yields coherence-based recovery guarantees and robustness (impossibility) certificates for $K$-flip budgets. Experiments on synthetic dictionaries and the Stanford Human Preferences dataset validate the theory and highlight how dictionary geometry governs attack success.

Theory · Reinforcement Learning and Planning

Debabrota Basu, Udvas Das, Brahim Driss, Uddalak Mukherjee

Post-deployment machine learning algorithms often influence the environments they act in, and thus *shift* the underlying dynamics that the standard reinforcement learning (RL) methods ignore. While designing optimal algorithms in this *performative* setting has recently been studied in supervised learning, the RL counterpart remains under-explored. In this paper, we prove the performative counterparts of the performance difference lemma and the policy gradient theorem in RL, and introduce the **Performative Policy Gradient** algorithm **PePG**. **PePG** is the first policy gradient algorithm designed to account for performativity in RL. Under softmax parametrisation, and also with and without entropy regularisation, we prove that **PePG** converges to *performatively optimal policies*, i.e. policies that remain optimal under the distribution shifts induced by themselves. Thus, **PePG** significantly extends the prior works in Performative RL that achieves *performative stability* but not optimality. Our empirical analysis on standard performative RL environments validate that **PePG** outperforms the existing performative RL algorithms aiming for stability.

Deep Learning · Generative Models and Autoencoders

Shanchuan Lin, Ceyuan Yang, Zhijie Lin, Hao Chen, Haoqi Fan

We present adversarial flow models, a class of generative models that belongs to both adversarial and flow families of models. Our method supports native one-step or multi-step generation and is trained using the adversarial objective. Unlike traditional GANs, where the generator learns an arbitrary transport plan between the noise and the data distributions, our generator learns a deterministic noise-to-data mapping, which is the same optimal transport as in flow-matching models. This significantly stabilizes adversarial training. Also, unlike consistency-based methods, our model directly learns one-step or few-step generation without having to learn the intermediate timesteps of the probability flow for propagation. This saves model capacity, reduces training iterations, and avoids error accumulation. Under the same 1NFE setting on ImageNet-256px, our B/2 model approaches the performance of consistency-based XL/2 models, while our XL/2 model creates a new best FID of 2.38. We additionally show the possibility of end-to-end training of 56-layer and 112-layer models through depth repetition without any intermediate supervision, and achieve FIDs of 2.08 and 1.94 using a single forward pass, surpassing their 28-layer 2NFE and 4NFE counterparts.

Raj Ghugare, Michał Bortkiewicz, Alicja Ziarko, Benjamin Eysenbach

How does the amount of compute available to a reinforcement learning (RL) policy affect its learning? Can policies using a fixed amount of parameters, still benefit from additional compute? The standard RL framework does not provide a language to answer these questions formally. Empirically, deep RL policies are often parameterized as neural networks with static architectures, conflating the amount of compute and the number of parameters. In this paper, we formalize compute bounded policies and prove that policies which use more compute can solve problems and generalize to longer-horizon tasks that are outside the scope of policies with less compute. Building on prior work in algorithmic learning and model-free planning, we propose a minimal architecture that can use a variable amount of compute. Our experiments complement our theory. On a set 31 different tasks spanning online and offline RL, we show that $(1)$ this architecture achieves stronger performance simply by using more compute, and $(2)$ stronger generalization on longer-horizon test tasks compared to standard feedforward networks or deep residual network using upto 5 times more parameters.