Long-Horizon Tasks

Many robotic tasks require short-term memory, whether it's retrieving an object that's no longer visible or turning off an appliance after a set period. Yet, most visuomotor policies trained via imitation learning rely only on immediate sensory input without using past experiences to guide decisions. We present PRISM, a transformer-based architecture for visuomotor policies to effectively use short-term memory via two key components: (i) gated attention, which filters retrieved information to suppress irrelevant details, improving performance by reducing the spurious correlations between the history and current action prediction, (ii) a hierarchical architecture that first compresses local information into compact tokens and then integrates them to capture temporally extended dependencies, improving its compute and memory footprint. Together, these mechanisms enable us to scale short-term memory in visuomotor policies for up to two minutes. To systematically evaluate memory in visuomotor control, we introduce ReMemBench -- a benchmark of eight diverse household manipulation tasks spanning four categories of short-term memory -- designed to foster general memory mechanisms rather than siloed, task-specific solutions. PRISM consistently outperforms prior works, including recurrent architectures, transformers, and their variants -- achieving an absolute improvement of 5%--12% over the strongest baseline. On the RoboCasa and LIBERO benchmarks, it achieves absolute improvements of 11%--15% over its no-memory variant and fine-tuned Vision-Language-Action baselines such as GR00T-N1-3B and OpenVLA, despite not leveraging any large-scale pretraining. Together, PRISM and ReMemBench establish a foundation for developing and evaluating short-term memory-augmented visuomotor policies that scale to long-horizon tasks. Additional materials are available at https://shahrutav.github.io/short-term-memory

LLM-based agents mark a shift from passive question answering to active task completion: they perceive environments, invoke tools, maintain state, and act over extended horizons. As agent systems have evolved from prompt engineering to workflows and context engineering, harness engineering, and agent-native training with co-evolution, a central question has become increasingly important: where does the bottleneck in agent performance reside, in the foundation model, in the execution harness, or in the coupling between them? This survey examines LLM-based agents through a model-harness lens. We first clarify the functional definition of agents and the implementation view of an LLM-based agent as a foundation model coupled with an execution harness. We then analyze the limits of model-centric scaling, trace four paradigms of agent engineering, and decompose the execution harness into six coupled runtime responsibilities: observation, context, control, action, state, and verification. Using this decomposition, we map task properties and domain pressures to harness configurations, review benchmark and evaluation practices, and synthesize model-harness evidence on how runtime design affects long-horizon task completion, efficiency, and reliability. Finally, we identify open challenges in value-aware evaluation, safety, harness generalization, and model-harness co-evolution. Rather than treating agents as models with auxiliary tools, this survey argues that agent quality -- including success, efficiency, safety, and generalization -- emerges from the interaction between model capability, runtime infrastructure, task structure, and evaluation design. A collection of papers discussed in this survey is provided in https://github.com/ggjy/Awesome-Agent-Engineering.

Existing computer-use agents remain fundamentally limited in professional software manipulation: GUI-based agents suffer from fragile visual grounding and long-horizon error accumulation, while API-basedapproaches struggle with heterogeneous protocols and inaccessible commercial interfaces. In this work,we identify the Component Object Model (COM) as a unified executable abstraction, proposing COM-as-Action: a new paradigm that reframes professional software interaction as deterministic program synthesisrather than sequential visual control. To validate this paradigm in the most demanding environments, weintroduce ComCADBench, the first benchmark for agents operating real industrial CAD software. Ourexperiments reveal a substantial paradigm gap: frontier proprietary models achieve near-zero successunder GUI-based interaction, whereas COM-based execution yields substantial immediate gains. Tobridge the remaining gap between syntactic correctness and geometric accuracy, we develop ComActor, aself-correcting agent trained through a progressive three-stage framework, alongside ComForge, a scalableplatform for large-scale training in Windows containers. Extensive experiments show that ComActorachieves state-of-the-art performance on ComCADBench, with strong resilience in long-horizon taskswhere baselines collapse, and generalizes to external CAD benchmark.

Recent years have witnessed the rapid evolution of AI agents toward handling increasingly complex, real-world tasks. However, existing benchmarks rarely evaluate whether agents can operate graphical user interfaces to complete long-horizon, high-value professional workflows across diverse domains. Current GUI benchmarks still predominantly focus on general-purpose software, relatively simple applications, and short-horizon tasks, leaving it largely unknown whether modern agents can follow user instructions to autonomously operate domain-specific professional software and accomplish economically valuable work in an end-to-end manner. To bridge this gap, we introduce Workflow-GYM, a benchmark for long-horizon GUI tasks centered on professional domains and specialized software environments. Through extensive experiments on state-of-the-art models, we find that even the strongest models achieve only slightly above 30% success rates, highlighting that professional long-horizon GUI workflows remain highly challenging for current GUI agents. Further analysis reveals that current agents struggle to maintain long-horizon workflow consistency, frequently exhibiting workflow stage omission, error propagation, objective drift, and insufficient understanding of professional software environments. Our findings provide important insights into the limitations of current agent systems and suggest key directions for the next generation of GUI-agent research.

Temporal modeling is essential for robotic manipulation, as effective control requires both memory of past interactions and imagination of future states. However, most VLA models rely primarily on the current observation and therefore struggle with long-horizon, temporally dependent tasks. Cognitive science suggests that humans rely on working memory to buffer short-lived context, the hippocampal system to preserve episodic memory of past experience, and internal models to imagine possible future state evolution. Inspired by these mechanisms, we propose MemoryVLA++, a full temporal modeling framework that equips VLA models with memory and imagination for robotic manipulation. A pretrained VLM encodes the current observation into perceptual and cognitive tokens, forming working memory. These tokens query a Perceptual-Cognitive Memory Bank to retrieve relevant historical context. This bank stores low-level details and high-level semantics from past interactions, and is updated through redundancy-aware consolidation. A world model imagines future states in a denoising latent space, and the imagined latents are integrated under memory guidance to form full temporal-aware tokens. The resulting tokens condition a diffusion action expert to predict temporally consistent action sequences. We conduct extensive experiments on 5 simulation benchmarks and 3 categories of real-robot tasks across 3 robots, covering general manipulation, long-horizon temporal tasks, robustness, and generalization. Our method achieves strong performance across Libero, SimplerEnv, Mikasa-Robo, Calvin, Libero-Plus, and diverse real-robot tasks, validating the effectiveness of full temporal modeling with memory and imagination. For example, on real robots, it achieves +9%, +26%, +28% gains on general, memory-dependent, and imagination-dependent tasks. Project Page: https://shihao1895.github.io/MemoryVLA-PP-Web

Vision-Language Models (VLMs) are increasingly required to process unbounded video streams in applications such as video-call assistants, live commentary, and embodied robots. An ideal streaming system should support proactive interaction, long-horizon memory, and real-time processing, while resting on a VLM backbone capable of handling diverse in-the-wild streaming tasks. However, existing VLMs excel at offline video understanding but fall short in streaming capabilities and lack dedicated infrastructure for streaming deployment. We address this gap on three fronts. (i) For backbone capability, we construct \textbf{Streaming-Train-248K}, a streaming dataset paired with a novel training objective for adapting VLMs to streaming interaction and understanding. (ii) For real-world deployment, we introduce \textbf{Streaming Harness}, a plug-and-play system that endows any VLM with three core abilities: proactive interaction (per-second response decisions), long-term memory (12-hour context retention), and real-time processing (sub-second latency). (iii) To drive continued community progress on streaming capabilities, we design \textbf{Streaming-Eval}, a benchmark that reflects models' capabilities across diverse in-the-wild scenarios. Extensive experiments demonstrate consistent gains from our approach across all core capabilities required for streaming video understanding. We will open-source our data, code, and benchmark to advance the community's shift from offline video understanding to deployable streaming intelligence.

AI agents are increasingly expected to complete long-horizon workflows that require sustained progress over hours, millions of tokens, and complex environments. Yet current agent benchmarks largely evaluate short-form tasks, such as single pull requests, small tickets, or 5-10 minute exercises, limiting our ability to measure agents' capabilities in planning, long-context understanding, and memory use. We introduce SWE-Marathon, a benchmark of 20 long-horizon tasks spanning software engineering and adjacent technical domains. Each task consists of a unique executable environment, a human-written reference solution, and a multi-layer verification suite. Logged agent attempts average 27.2M total tokens, making SWE-Marathon substantially longer-horizon than existing SWE and command-line agent benchmarks. Current frontier coding agents solve fewer than 30% of tasks. Failures often arise from poor self-verification, self-reported infeasibility, and premature termination. We also observe reward-hacking behavior in 13.8% of rollouts, where agents attempt to exploit the environment or verifier to bypass the intended workflow. SWE-Marathon includes adversarial review of test suites and execution environments, as well as multi-layer checks designed to prevent shortcut solutions. We release SWE-Marathon, evaluation code, and agent trajectories at https://swe-marathon.org/.

LLM-based agents increasingly tackle long-horizon tasks with interdependent decisions, where each action reshapes future constraints and intermediate errors can cascade. Existing RAG and agent memory systems organize histories by semantic similarity, retrieving content-relevant entries at decision time. We argue that this design mismatches execution-state dependencies: it fragments decision trajectories and mixes valid and erroneous traces, hindering coherent state reconstruction and error isolation. We propose MAGE (Memory as Agent-Guided Exploration), an active execution-state manager that stores interactions in a hierarchical state tree. The agent derives its state from the active root-to-current path, combining subgoal summaries, recent traces, and hints from prior branches. Four coupled operations maintain the tree: Grow records new traces, Compress summarizes completed subgoals, Maintain validates summaries, and Revise restores a target boundary and resumes on a new branch. This design bounds context growth while preserving state integrity and isolating flawed segments from the active path. Experiments on MemoryArena show that MAGE improves the average task success rate by 7.8--20.4 pp over baselines, while reducing token consumption by 55.1%.

Despite recent progress, LLM agents still struggle with reasoning over long interaction histories. While current memory-augmented agents rely on a static retrieve-then-reason paradigm, this rigid pipeline design prevents them from dynamically adapting memory access to intermediate evidence discovered during inference. To bridge this gap, we propose MRAgent, a framework that combines an associative memory graph with an active reconstruction mechanism. We represent memory as a Cue-Tag-Content graph, where associative tags serve as semantic bridges connecting fine-grained cues to memory contents. Operating on this structure, our active reconstruction mechanism integrates LLM reasoning directly into memory access, allowing the agent to iteratively explore and prune retrieval paths based on accumulated evidence. This ensures that memory retrieval is dynamically adapted to the reasoning context while avoiding combinatorial explosion caused by unconstrained expansion. Experiments on the LoCoMo benchmark and LongMemEval benchmark demonstrate significant improvements over strong baselines (up to 23%), while substantially reducing token and runtime cost, highlighting the effectiveness of active and associative reconstruction for long-horizon memory reasoning.

We propose world-language-action (WLA) models as a new class of embodied foundation models. WLA takes textual instructions, images, and robot states as inputs to jointly predict textual subtasks, subgoal images, and robot actions, conjoining the \emph{world modeling interface} to learn from extensive egocentric videos as in the world-action model (WAM) and the \emph{language reasoning} capacities to solve complex long-horizon tasks as in vision-language-action (VLA) models. At the core of WLA lies an \emph{autoregressive (AR)} Transformer backbone, instead of a bidirectional diffusion Transformer as in WAMs, to predict the \emph{next state}, comprising the \emph{semantic-level} textual intention and complementary \emph{fine-grained} physical dynamics. The physical dynamics are supervised by the world modeling objective based on a dedicated World Expert, and are leveraged to ease the characterization of the state-action correlation for the Action Expert. WLA leverages meta-queries to make the world prediction \emph{implicitly} impact the action generation so that the former can be disabled during inference. The world prediction can also be activated to enable test-time scaling for improved robot control. Our WLA-0 prototype, with 2B active parameters, achieves 40 ms per inference on an NVIDIA RTX 5090. Evaluations across simulated and real-world environments demonstrate that WLA-0 achieves state-of-the-art multi-task and long-horizon learning abilities, e.g., 92.94% success rate on RoboTwin2.0 Clean and 56.5% success rate on RMBench. WLA-0 also holds the promise to learn novel tasks directly from \emph{cross-embodiment robot videos} without action annotations.

Embodied AI systems are increasingly expected to reason and act over extended horizons in physical environments. This growing capability brings safety to the foreground, because failures in the physical world can harm people, damage objects, and disrupt workplaces. Although safe embodied AI has attracted substantial attention, the literature remains fragmented across planning, policy design, and runtime execution. Long-horizon robotic manipulation is a particularly revealing anchor domain for this problem because semantic misgrounding, subtask-level error propagation, execution drift, and contact-rich physical risk can accumulate within the same closed-loop system. This survey therefore provides a structured review of safety in long-horizon robotic manipulation from an embodied AI perspective. We organize the literature by intervention locus, covering planning-time, policy-time, and execution-time safety, and we analyze the strength of the evidence that each line of work provides, distinguishing formal guarantees, statistical support, and empirical safety heuristics. This framework clarifies the distinct roles of backbone capability papers, direct safety mechanisms, and benchmark or evaluation studies, while exposing where current safety claims are well supported and where they remain indirect. We identify persistent gaps, including limited evidence for policy-time safety, weak formal support for contact-rich long-horizon manipulation, immature uncertainty-triggered intervention, and a shortage of manipulation-specific safety benchmarks. We conclude by outlining research directions for cross-layer assurance, evaluation design, and safer deployment of long-horizon robotic agents in real-world settings.

Scientific and engineering progress is fundamentally a long-horizon iterative process: proposing changes, running experiments, measuring outcomes, and continuously refining artifacts. Yet existing benchmarks for frontier models primarily evaluate either single-turn responses or short-horizon agent trajectories, failing to capture the challenges of sustained iterative improvement over extended time horizons. To address this gap, we introduce AutoLab, a new benchmark for ultra long-horizon closed-loop optimization. AutoLab consists of 36 realistic, expert-curated tasks spanning four diverse domains: system optimization, puzzle & challenge, model development, and CUDA kernel optimization. Each task begins with a correct but deliberately suboptimal baseline and challenges agents to improve it within a strict wall-clock budget. Evaluating 17 state-of-the-art models reveals the dominant predictor of success is not the quality of an agent's initial attempt, but its persistence in repeatedly benchmarking, editing, and incorporating empirical feedback. While claude-opus-4.6 exhibits strong long-horizon optimization capabilities, most frontier models, including several proprietary ones, either terminate prematurely or exhaust their budgets with minimal progress. These results underscore the importance of time awareness and persistent iteration in autonomous agents. We open-source the full benchmark, evaluation harness, and task artifacts, to accelerate research toward truly capable long-horizon agents.

Recent progress in Large Language Model (LLM) agents has enabled promising advances in automated data science. However, existing approaches remain fundamentally limited by their static action sets and lack of principled long-horizon context management, hindering their ability to accumulate reusable experience across tasks and operate reliably in multi-stage, iterative data science pipelines. To address these challenges, we introduce EvoDS, a self-evolving autonomous data science agent that learns to expand its skills and adaptively managing long-term context through agentic reinforcement learning. Specifically, EvoDS introduces two key strategies: (1) Autonomous Skill Acquisition (ASA) mechanism, which enables agents to synthesize, validate, and reuse executable skills; and (2) Adaptive Context Compression (ACC) strategy, which treats context management as a learned control problem rather than passive truncation. These strategies are orchestrated within a two-stage multi-agent training scheme, enabling EvoDS to autonomously improve over time. Theoretically, we prove that EvoDS's hierarchical design reduces tool-selection error, and its optimization objective aligns with an information bottleneck principle, ensuring efficient context use. Empirically, EvoDS outperforms state-of-the-art open-source data science agents by an average of 28.9% across four diverse benchmarks while eliminating out-of-token failures. Our code and data are available at https://github.com/usail-hkust/EvoDS.

Real-world professional desktop workflows in specialized creative and engineering software unfold over long horizons and often require human-in-the-loop coordination, where agents proactively seek necessary information and users provide additional instructions, clarifications, feedback, or corrections as the task progresses. Yet existing desktop GUI benchmarks mostly reduce this setting to short, simplified tasks with all user instructions provided upfront. To address this issue, we introduce DeskCraft, a desktop GUI benchmark targeting long horizon creative and engineering workflows and proactive human-agent collaboration. DeskCraft organizes tasks into a multilevel difficulty taxonomy, with long horizon tasks requiring over 50 execution steps, and covers professional creative software across design, video, audio, and 3D creation. Furthermore, DeskCraft formalizes human-agent collaboration into an interaction protocol covering mid-turn and post-turn exchanges. Mid-turn interaction captures both agent-initiated clarification under uncertainty and user-initiated interruption during execution, while post-turn interaction accommodates user-driven feedback after the agent signals completion, together spanning the full space of realistic collaboration patterns. We evaluate 18 proprietary and open source agents on 538 tasks and find that GPT-5.4 reaches 31.6% on standard tasks and 27.6% on interactive tasks. Further analyses reveal persistent failures in long horizon workflow delivery and proactive clarification. We will open-source all evaluation codes, tasks, and data at https://github.com/mrwwk/DeskCraft.

Computer use agents (CUAs) today are primarily deployed as single serial agents. This setup is suboptimal for complex long-horizon tasks that benefit from task decomposition, parallel execution, and consistent re-planning based on new information. In this paper, we argue that we should instead move towards evaluating and building multi-agent computer use (MACU) systems. These systems, which emphasize planning and parallel execution, alleviate many of the shortcomings of single-agent CUAs. We propose a general multi-agent setup in which a manager model decomposes computer use tasks as a directed acyclic graph (DAG), encoding relevant dependencies and goals for subagents. At each iteration, the manager dispatches parallel CUA subagents to carry out nodes on the ready frontier of the DAG, and continuously revises the DAG (adding, canceling, or rewriting nodes) as new findings arrive from subagents. This design treats the partially observable environment of computer use as a first class challenge: information that downstream agents may not be able to re-observe are retained and passed forward through the manager and DAG structure. We demonstrate that MACU consistently improves over strong single-agent baselines by $3.4-25.5\%$ on desktop (OSWorld) and web navigation (Online-Mind2Web, WebTailBench, Odysseys) benchmarks, exhibits more favorable test-time scaling, and solves complex long-horizon tasks where single-agent CUAs get stuck. On Odysseys, a long-horizon web navigation benchmark, MACU improves average task completion wall-clock time by ${\sim} 1.5 \times$, demonstrating its efficacy in speeding up traditionally slow CUA pipelines. Our findings highlight that multi-agent coordination is a promising axis for scaling computer use agents to work productively for longer and more effectively. We release all code and interactive visualizations at https://jykoh.com/multi-agent-computer-use.

Large language models are increasingly deployed as tool-augmented agents to acquire information beyond parametric knowledge. While recent work has improved long-horizon tool-use reasoning, most approaches focus on tasks with a single correct answer. In contrast, many real-world queries require discovering a comprehensive set of valid answers, a setting known as Multi-Answer QA. This setting raises two challenges: fine-grained credit assignment over long search trajectories and reward alignment for sustained exploration beyond easy high-frequency entities. We propose SPADER, a reinforcement learning framework for long-horizon tool use in Multi-Answer QA. SPADER includes Step-wise Peer Advantage (SPA), a critic-free step-level credit assignment mechanism that aligns parallel trajectories by decision step and estimates advantages from peer returns. It also includes a diversity-aware exploration reward that promotes long-tail entity discovery by upweighting rare findings and downweighting redundant ones. Experiments on QAMPARI, Mintaka, WebQSP, and QUEST show that SPADER generally improves recall and overall F1 over prompting-based agents, outcome-supervised RL methods, and recent step-level supervision approaches. Our code and model weights are available at https://github.com/KhanCold/spader.

LLM agents increasingly face long-horizon tasks such as web search and deep research in real-world applications, where accumulated context can cause long-context degradation and reasoning failures. Prior work mitigates this through context management with agent-side context control or fixed strategies such as summarization, which require training the agent itself for adaptation - making it impractical for closed-source agents and ignoring that different agents may require different strategies. We introduce Adaptive Context Management (AdaCoM), which trains an external LLM to manage the context of a frozen agent through flexible modification actions and end-to-end reinforcement learning. Across diverse agents on web search and deep research benchmarks, AdaCoM substantially improves performance by preserving task constraints and progress while pruning stale content. The learned strategies reveal a Fidelity-Reliability Trade-off: agents with higher vanilla ReAct performance benefit from higher-fidelity context preservation, whereas lower-performing agents require more aggressive compression to stay within a reliable reasoning regime. Transfer experiments show that AdaCoM generalizes most effectively across agents with similar capability (measured by vanilla ReAct performance), suggesting a practical path toward reusable context managers for agent systems.

Web agents, which couple language models with browsing and tool-use capabilities, show promise as open web assistants. Yet progress is increasingly limited by the lack of scalable, process-level supervision. Existing benchmarks are largely manually constructed, providing only coarse start-goal annotations without intermediate trajectories, while recent automatic generation efforts remain expensive, biased, and shallow. These limitations prevent reliable training and evaluation of agents that must generalize to realistic, multi-hop, cross-page tasks. We introduce a scalable framework, GTA, that integrates crawling, retrieval-based seeding, in-context generation, and automated quality control to produce realistic tasks paired with executable trajectories. This design decouples crawling from generation for greater efficiency, grounds tasks in the site graph to enforce compositionality, and ensures dense supervision through deterministic replays and systematic validation. We instantiate the pipeline on over 50 websites covering e-commerce, government, forums, and news, with multilingual and multi-hop coverage. The resulting benchmark reveals a significant human-agent performance gap and enables detailed diagnostics. Our contributions are three-fold: (i) formalizing multi-hop web-agent task generation, (ii) proposing an efficient and validated pipeline for automatic data creation, and (iii) releasing a dynamic benchmark with reproducible evaluation.

Offline goal-conditioned reinforcement learning (GCRL) is challenging in long-horizon tasks, where distant state--goal pairs provide weak supervision and value estimates become vulnerable to accumulated bootstrapping errors. Hierarchical methods mitigate this difficulty by introducing intermediate subgoals, but fixed temporal abstractions or fixed hierarchy depths can be mismatched to state--goal pairs with different reachability horizons. We propose Coarse-to-Fine Hierarchical Goal Reinforcement Learning (CFHRL), a fully offline GCRL framework that adaptively refines distant goals before execution. Starting from the final goal, CFHRL recursively proposes intermediate targets, trained from replay-supported candidates, and stops refinement once the current target is estimated to be locally executable by a learned reachability cost. The key idea is that a subgoal need not be an exact midpoint or globally optimal waypoint; it only needs to provide reliable progress and reduce the remaining reaching difficulty, enabling subsequent refinement over shorter horizons. A stylized analysis further supports the robustness of approximate recursive contraction. Experiments on OGBench show substantial gains on several long-horizon tasks, with ablations validating the proposed refinement and stopping mechanisms

We present PEAM, a Parametric Embodied Agent Memory framework in Minecraft that transforms agent memory from inference-time retrieval into parameter-resident skills internalized through experience. PEAM pairs a slow deliberative LLM for open-ended reasoning with a fast parametric module for reflexive execution of consolidated skills. The fast module is a multimodal Mixture-of-Experts LoRA architecture with per-category physically isolated adapters, enabling parameter-level continual learning without catastrophic forgetting. We treat failure as a first-class training signal: failure--correction trajectory pairs are internalized through a joint behavioral-cloning and contrastive objective, so the agent learns not only what succeeds but also how corrected actions differ from failed ones. To govern consolidation, PEAM introduces a parameterization-worthiness score for deciding which experience should be internalized, and a scale-free self-triggered consolidation mechanism for deciding when to internalize without task-specific hand-tuned thresholds, making the agent self-evolving as the trigger transfers across task distributions without re-tuning. Experiments in Minecraft show that PEAM improves long-horizon task performance, mitigates forgetting on previously consolidated skills, and improves parametric-versus-retrieval efficiency over retrieval-based embodied agents and parametric memory variants.

AI agents are beginning to complete valuable, long-horizon business operations tasks, but training and evaluation environments for enterprise work still struggle to balance realism, verifiability, and scale. Environment and task creation frequently suffers from a failure mode we call artifact drift: when instructions, environments, oracles, and verifiers are created by loosely coupled processes, they frequently disagree on what a task requires, producing environments that are unsolvable, reward-hackable, or inconsistent. We introduce Anchor, a task-generation pipeline that formalizes domain experts' specifications of business workflows into constraint optimization programs. From a single parametric specification, the pipeline jointly produces a natural-language instruction, environment configuration, solver-certified ground-truth solution, and state-based verifier. With Anchor, altering parameters yields new tasks with controlled difficulty and known optimal solutions, producing harness-agnostic environments whose rewards depend solely on end-state business correctness. We apply Anchor to produce ERP-Bench: a benchmark of 300 long-horizon tasks spanning procurement and manufacturing workflows in a production-grade ERP system. We find that generation parameters predict realized difficulty, and that frontier models satisfy explicit task constraints in 26.1% of trials but reach a fully optimal solution in only 17.4% of trials. Overall, we show that Anchor and ERP-Bench offer a concrete recipe for building auditable evaluation environments for economically valuable agent work. We release the task generator and ERP-Bench dataset at erpbench.ai

Offline goal-conditioned reinforcement learning (GCRL) provides a practical framework for obtaining goal-reaching policies from fixed datasets. However, learning a reliable goal-conditioned value function in long-horizon tasks remains challenging. In this paper, we identify erroneous generalization in goal-conditioned value functions as a fundamental bottleneck, and demonstrate that appropriate inductive bias in the value function is crucial for addressing the bottleneck. Building on these findings, we propose Latent-Aligned Value Learning (LAVL), an offline GCRL algorithm that integrates latent-representation-based value generalization with hierarchical planning in a unified framework. Extensive experiments on OGBench demonstrate that LAVL consistently outperforms existing offline GCRL methods, achieving the highest performance on 20 out of 22 datasets. Notably, LAVL exhibits strong performance in long-horizon tasks and trajectory stitching datasets, where prior methods suffer significant performance degradation. Our code is available at https://github.com/oh-lab/LAVL.git.

Large language model (LLM) agents excel at solving complex long-horizon tasks through autonomous interaction with environments. However, their real-world deployment faces a fundamental device--cloud dilemma: on-device models are efficient but often brittle, while cloud models are stronger but costly in computation. State-of-the-art LLM device--cloud routers usually make coarse task-level decisions, which cannot adapt to the changing difficulty of multi-step agent interactions. To address this issue, we present Hera, a step-level device--cloud LLM agent coordinator for long-horizon tasks achieving a strong performance--cost Pareto frontier. Hera adopts a novel two-stage training paradigm: (1) imitation learning for cold-start, followed by (2) reinforcement learning that jointly optimizes task success and cloud usage efficiency. The first stage casts step-level routing as a supervised classification problem: the device agent is replayed on cloud trajectories, with each state labeled by the agreement between device and cloud actions. In the second stage, we perform cost-aware reinforcement learning by grouping identical states across trajectories and updating Hera with labels favoring higher expected return and fewer future cloud calls. We evaluate Hera on ALFWorld, WebShop, and AppWorld, where it consistently outperforms prior methods, achieving 92.5% of the cloud-only success rate with cloud use in only 46.3% of steps.

Recent progress on long-horizon agentic tasks has been driven largely by scaling up individual agents through stronger models, better tools, and more effective scaffolding. In contrast, much less is understood about scaling out: whether multiple peer agents, all targeting the same task, can become an additional source of capability without relying on explicit role specialization or workflow orchestration. We study this question and propose AgentFugue, a collective reasoning framework built around a shared reasoning hub. As peer agents explore the same task in parallel, the hub records concise notes on what each agent has established, attempted, or ruled out, and enables each agent to selectively access what other agents have discovered in a form useful for its current search. This design turns otherwise isolated trajectories into a connected ecology of reusable intermediate reasoning without requiring centralized planning. We instantiate the hub as a plug-in communication layer, trained with supervised fine-tuning and end-to-end reinforcement learning. Across the challenging long-horizon settings we study, AgentFugue improves over strong baselines. Our results suggest that collective reasoning can turn scaling out peer agent systems into a distinct source of capability gains, rather than merely a way of spending more compute.

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 to 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 research agents extend the role of search engines from retrieving keyword-matched pages to synthesizing knowledge, fundamentally changing how humans interact with information. However, frontier systems remain proprietary, while existing open agents often generalize poorly across different task types, leaving unclear how to train a broadly capable deep research agent. We release QUEST, a family of open models (ranging from 2B to 35B) that serve as general-purpose deep research agents designed to handle a wide range of long-horizon search tasks, with strong capabilities in fact seeking, citation grounding, and report synthesis. To build QUEST, we propose an effective training recipe combining mid-training, supervised fine-tuning, and reinforcement learning. Central to this recipe is a curated data synthesis pipeline based on unified rubric trees, which applies to different task types and enables synthesizing training data with verifiable rewards without human annotation. In addition, QUEST incorporates a built-in context management mechanism that enables effective long-horizon reasoning and knowledge synthesis. Using only 8K synthesized tasks, QUEST approaches or even surpasses frontier closed-source agents across eight deep research benchmarks spanning diverse task types, and achieves the best overall performance among recent open-weight agents. We released everything: models, data, and training scripts.

Large language model (LLM)-based agents solve complex tasks by leveraging multi-step reasoning with iterative tool calls and environment interactions, which incur idle time while waiting for observations. Despite the prevalence of idle time in most agentic scenarios, existing works treat it as an unavoidable overhead or propose restricted solutions that overlook varying computational budgets across different tool calls and future observation uncertainty, thereby leading to suboptimal utilization of idle time. In this paper, we introduce IdleSpec, a scalable and generic inference approach that leverages idle-time computation to improve agent performance while minimizing latency overhead. Specifically, IdleSpec iteratively generates plan candidates during idle periods and, once observations become available, aggregates them to guide the next reasoning step. For effective plan generation under observation uncertainty, IdleSpec samples between complementary drafting strategies (i.e., progressive and recovery) from a learned distribution that is updated via posterior feedback. Our experiments demonstrate that IdleSpec significantly improves agent performance in various agentic scenarios by effectively utilizing idle time. In particular, on the GAIA and FRAMES, IdleSpec achieves 55.6% average accuracy with Gemini-2.5-Flash, surpassing the vanilla baseline without idle-time usage by 5.1%. Furthermore, for MLE-Bench, which involves substantial delay from code executions, IdleSpec achieves performance gains of up to 9.1% on the Any Medal rate, highlighting its generalizability to long-horizon tasks.

Real-world agents operate over long and evolving horizons, where information is repeatedly updated and may interfere across memories, requiring accurate recall and aggregated reasoning over multiple pieces of information. However, existing benchmarks focus on static, independent recall and fail to capture these dynamic interactions between evolving memories. In this paper, we study how current memory-augmented agents perform in realistic, interference-heavy, long-horizon settings across diverse domains and question types. We introduce MINTEval (Long-Horizon Memory under INTerference Evaluation), a benchmark featuring (1) long, highly interconnected contexts with frequently updated information that induces substantial interference, (2) diverse domains (state tracking, multi-turn dialogue, Wikipedia revisions, and GitHub commits), enabling evaluation of domain generalization, and (3) diverse question types that assess robustness to interference, including (i) single-target recall tasks requiring retrieval of a specific target from long contexts, and (ii) multi-target aggregation tasks requiring reasoning over multiple relevant pieces of information. Overall, MINTEval has 15.6k question-answering pairs over long-horizon contexts averaging 138.8k tokens and extending up to 1.8M tokens per instance. We evaluate 7 representative systems, including vanilla long-context LLMs, RAG, and memory-augmented agent frameworks. Across all systems, we observe consistently low performance (avg. 27.9% accuracy), especially on questions requiring aggregated reasoning over multiple pieces of evidence. Our analysis shows that performance is primarily limited by retrieval and memory construction. Furthermore, current memory systems struggle to recall and reason over earlier facts that are revised or interfered with by subsequent context, with accuracy degrading as the number of intervening updates increases.

As Large Language Models (LLMs) evolve into persistent scientific collaborators, context window saturation has emerged as a critical bottleneck. Scientific workflows involving iterative data analysis and hypothesis refinement rapidly saturate even extended contexts with dense technical content, while monolithic approaches suffer from quadratic cost scaling and cognitive degradation. We evaluate a Dual Process Memory Architecture that decouples immediate episodic needs (constant 10-message window) from long-term consolidated knowledge (growing at approximately 3 tokens/message). Unlike prior social agent memory systems, our domain-specific consolidation addresses contradictory parameter evolution, multi-hop reasoning across experimental phases, and precise technical fact retention. Through large-scale evaluation spanning 15,000 messages with cross-model validation across six LLMs from three families (OpenAI, Anthropic, Google), totaling 1,440 queries, we establish three key findings. First, while full-context models fail at 10,000 messages due to context overflow, our system maintains 70-85% accuracy with 1-2 second latency using 62% fewer tokens (45,434 vs 120,000+ limit). Second, cross-model validation reveals architecture-level trade-offs independent of specific LLMs: Dual Process excels at numeric/temporal queries (65-90% accuracy) while RAG excels at historical retrieval (60-85%), suggesting complementary deployment strategies. Third, we identify a "Sim-to-Real" gap where synthetic tests maintain constant memory but realistic workflows exhibit linear growth (about 3 tokens/message), with consolidation quality emerging as the primary scalability bottleneck. The architecture successfully manages profiles with 14,000+ scientific facts (125k tokens), demonstrating that domain-specific memory consolidation enables sustained operation beyond full-context limits.

As autonomous coding agents become capable of handling increasingly long-horizon tasks, they have gradually demonstrated the potential to complete end-to-end software development. Although existing benchmarks have recently evolved from localized code editing to from-scratch project generation, they remain confined to structurally simplified, single-stack applications. Consequently, they fail to capture the heterogeneous environments, full-stack orchestration, and system-level complexity of real enterprise Software as a Service (SaaS) systems, leaving a critical gap in assessing agents under realistic engineering constraints. To fill this gap, we introduce SaaSBench, the first benchmark designed to explore the boundaries of AI agents in enterprise SaaS engineering. Spanning 30 complex tasks across 6 SaaS domains with 5,370 validation nodes, it incorporates 8 programming languages, 6 databases, and 13 frameworks to meticulously mirror real-world software heterogeneity. Furthermore, we design a dependency-aware hybrid evaluation paradigm tailored for complex systems with long horizons and multi-component coupling, enabling fine-grained, reproducible assessment. Crucially, our extensive experiments reveal a striking insight: the primary bottleneck for state-of-the-art agents is not generating isolated code logic, but successfully configuring and integrating a multi-component system. Over 95% of task failures occur before agents even reach deep business logic, with models often falling victim to overconfidence and prematurely halting during foundational system setup, or getting trapped in ineffective debugging loops. We hope SaaSBench serves as a practical and challenging testbed to drive the evolution of reliable, system-level coding agents. The code is available at \url{https://github.com/ShadeCloak/SaaSbench}.

Learning visuomotor policies from scarce expert demonstrations remains a core challenge in robotic manipulation. A primary hurdle lies in distilling high-dimensional RGB representations into control-relevant geometry without overfitting. While using frozen pre-trained Vision Foundation Models (VFMs) improves data efficiency, it also shifts most task adaptation onto a small spatial pooling module, which can latch onto task-irrelevant shortcuts and lose geometric grounding when finetuned with few data samples. More broadly, pre-trained visual representations used for policy learning have been observed to struggle under even minor scene perturbations, highlighting the need for robustness-oriented inductive biases. We propose Geometric Anchor Pre-training (GAP), a simple, action-free warm-up stage that regularizes the spatial adapter before downstream imitation learning. GAP pre-trains the pooling layer on a lightweight simulated proxy task where object masks are available at no cost, encouraging the adapter to produce keypoints that lie on the object, cover its spatial extent, and remain sharp and repeatable over time. This yields stable geometric anchors that provide a reliable coordinate interface for few-shot policy learning, while keeping the VFM frozen. We evaluate GAP on RoboMimic and ManiSkill under severe data scarcity (15-50 demonstrations) and domain shift. A simple adapter regularized with GAP consistently outperforms stronger attention-based poolers and end-to-end fine-tuning, achieving 62% success on RoboMimic Can with 15 demonstrations (+16% over AFA), 63% on the long-horizon high-precision Tool Hang task with 50 demonstrations, and 61% on ManiSkill StackCube with 30 demonstrations (+11% over full fine-tuning). The proxy stage is lightweight and fully decoupled from downstream tasks, making it practical to reuse across environments and manipulation skills.

Autonomous language-model agents are increasingly evaluated on long-horizon tool-use tasks, but existing benchmarks rarely capture the complexity and nuance of real scientific work. To address this gap, we introduce Collider-Bench, a benchmark for evaluating whether LLM agents can reproduce experimental analyses from the Large Hadron Collider (LHC) using only public papers and open scientific software. Such analyses are often difficult to reproduce because the public toolchain only approximates the software used internally by the experimental collaborations, while the published papers inevitably omit implementation details needed for a faithful reconstruction. Agents must therefore rely on physical reasoning, domain knowledge, and trial-and-error to fill these gaps. Each task requires the agent to turn a published analysis into an executable simulation-and-selection pipeline and submit predicted collision event yields in specified signal regions. These predictions are evaluated with standard histogram metrics that provide continuous fidelity scores without a hand-written rubric. We also report the computational cost incurred by each agent per task. Finally, we evaluate the codebase and full session trace using an LLM judge to catch qualitative failure modes such as fabrications, hallucinations and duplications. We release an initial set of tasks drawn from LHC searches, together with a containerized sandbox and event simulation tools. We evaluate across a capability ladder of general purpose coding agents. Our results show that on average no agent reliably beats the physicist-in-the-loop solution.

Long-horizon language agents accumulate conversation history far faster than any fixed context window can hold, making memory management critical to both answer accuracy and serving cost. Existing approaches either expand the context window without addressing what is retrieved, perform heavy ingestion-time fact extraction at substantial token cost, or rely on heuristic graph traversal that leaves both accuracy and efficiency on the table. We present PRISM, a training-free retrieval-side framework that treats long-horizon memory as a joint retrieval-and-compression problem over a graph-structured memory. PRISM combines four orthogonal inference-time components: Hierarchical Bundle Search over typed relation paths, Query-Sensitive Edge Costing that aligns traversal with detected query intent, Evidence Compression that compresses the candidate bundle into a compact answer-side context, and Adaptive Intent Routing that routes most queries through zero-LLM tiers. By formulating retrieval as min-cost selection over typed path templates and pairing it with an LLM-side compression step, PRISM surfaces the right evidence under a strict context budget without any fine-tuning or modification to the upstream ingestion pipeline. Experiments on the LoCoMo benchmark show that PRISM delivers substantially higher LLM-judge accuracy than every same-protocol baseline at an order-of-magnitude smaller context budget, occupying a previously empty corner of the accuracy-context-cost frontier and demonstrating a superior balance between answer quality and retrieval efficiency.

Long video question answering requires locating sparse, time-scattered visual evidence within highly redundant content. Although current MLLMs perform well on short videos, long videos introduce long-horizon search and verification, which often necessitates multi-turn, agentic interaction. We show that existing LVU agents can exhibit "evidence misalignment": they produce correct answers that are not supported by the retrieved or inspected evidence. To characterize this failure, we introduce two diagnostics (temporal groundedness and semantic groundedness) and use them to reveal two pressures that amplify misalignment: prompt pressure from shared-context saturation at inference time and reward pressure from outcome-only optimization during training. These findings point to a structural root cause: the coupled agent paradigm conflates long-horizon planning with answer authority. We therefore propose the decoupled planner-inspector framework, which separates planning from answer authority and gates final answering on pixel-level verification. Across four long-video benchmarks, our framework improves both answer accuracy and evidence alignment, achieving 55.1% on LVBench and 62.0% on LongVideoBench while producing interpretable search trajectories. Moreover, the decoupled architecture scales consistently with increased search budgets and supports plug-and-play upgrades of the MLLM backbone without retraining the planner. Code and models are available at https://github.com/Echochef/VideoSEAL.

Solving complex long-horizon tasks requires strong planning and reasoning capabilities. Although datasets with explicit chain-of-thought (CoT) rationales can substantially benefit learning, they are costly to obtain. To address this challenge, we propose Hindsight Hint Distillation (HHD), which only requires easy-to-obtain question-answer pairs without CoT annotations. Inspired by how human teachers use student mistakes to provide targeted guidance, HHD synthesizes hindsight hints from the model's own failed self-rollouts and uses them to scaffold on-policy rollouts that successfully complete the tasks. The model then self-distills these scaffolded trajectories and generalizes to new problems without hint guidance. Experiments show that HHD significantly outperforms iterative RFT and trajectory-synthesis baselines, achieving an absolute improvement of 8% on SWE-bench Verified, while all baselines improve by only around 2%. Notably, the reasoning strategies induced by HHD generalize effectively to out-of-distribution tasks, yielding the largest gains on SWE-bench Multilingual despite no training on multilingual data. These results demonstrate that HHD can effectively synthesize expert-like reasoning from CoT-free data and substantially improve long-horizon performance.

Large language models (LLMs) are increasingly deployed on long-horizon tasks in partially observable environments, where they must act while inferring and tracking a complex environment state over many steps. This leads to two challenges: partial observability requires maintaining uncertainty over unobserved world attributes, and long interaction history causes context to grow without bound, diluting task-relevant information. A principled solution to both challenges is a belief state: a posterior distribution over environment states given past observations and actions, which compactly encodes history for decision making regardless of episode length. In LLM agents, however, the open-ended nature of text makes it unclear how to represent such a distribution. Therefore, we introduce Agent-BRACE: Agent Belief state Representation via Abstraction and Confidence Estimation, a method that decouples an LLM agent into a belief state model and a policy model, jointly optimized via reinforcement learning. The belief state model produces a structured approximation of the belief distribution: a set of atomic natural language claims about the environment, each annotated with an ordinal verbalized certainty label ranging from certain to unknown. The policy model conditions on this compact, structured approximate belief rather than the full history, learning to select actions under explicit uncertainty. Across long-horizon, partially observable embodied language environments, Agent-BRACE achieves an average absolute improvement of +14.5% (Qwen2.5-3B-Instruct) and +5.3% (Qwen3-4B-Instruct), outperforming strong RL baselines while maintaining a near-constant context window independent of episode length. Further analysis shows that the learned belief becomes increasingly calibrated over the course of an episode as evidence accumulates.

Large language and vision-language models increasingly power agents that act on a user's behalf through command-line interface (CLI) harnesses. However, most agent benchmarks still rely on synthetic sandboxes, short-horizon tasks, mock-service APIs, and final-answer checks, leaving open whether agents can complete realistic long-horizon work in the runtimes where they are deployed. This work presents WildClawBench, a native-runtime benchmark of 60 human-authored, bilingual, multimodal tasks spanning six thematic categories. Each task averages roughly 8 minutes of wall-clock time and over 20 tool calls, and runs inside a reproducible Docker container hosting an actual CLI agent harness (OpenClaw, Claude Code, Codex, or Hermes Agent) with access to real tools rather than mock services. Grading is hybrid, combining deterministic rule-based checks, environment-state auditing of side effects, and an LLM/VLM judge for semantic verification. Across 19 frontier models, the best, Claude Opus 4.7, reaches only 62.2% overall under OpenClaw, while every other model stays below 60%, and switching harness alone shifts a single model by up to 18 points. These results show that long-horizon, native-runtime agent evaluation remains a far-from-resolved task for current frontier models. We release the tasks, code, and containerized tooling to support reproducible evaluation.

LLM-based foundation agents that perceive, reason, and act across thousands of reasoning steps are rapidly becoming the dominant paradigm for deploying artificial intelligence in open-ended, long-horizon complex tasks. Despite this significance, the field remains overwhelmingly engineering-driven. Engineering practice has converged on useful primitives (tool loops, memory banks, harnesses, reflection steps), yet these are assembled by empirical trial and error rather than from first principles. Fundamental questions remain open: under what conditions does a long-running agent remain on-task? How should an agent respond when its environment exceeds its representational capacity? What architectural properties are necessary for safe self-improvement? We argue that cybernetics, the mid-twentieth-century science of control and communication in complex systems, provides the missing theoretical scaffold for foundation agents. By mapping six canonical laws of classical cybernetics onto six agent design principles, and synthesizing those principles into three engineering desiderata (reliability, lifelong running, and self-Improvement), we arrive at a framework termed Agent Cybernetics. Three application domains, code generation, computer use and automated research, exemplify the analytical framework of agent cybernetics by identifying failure modes and concrete engineering recommendations. We hope that agent cybernetics opens a new research venue and establishes the scientific foundation that foundation agents need for principled, reliable real-world deployment.

Autonomous robotic assembly of interlocking bricks demands seamless integration of long-horizon task reasoning, spatial grounding, and fine-grained manipulation. This paper presents BrickCraft, a compositional framework designed for long-horizon and generalizable interlocking brick assembly. BrickCraft models the assembly process using a relative formulation, where each step is anchored to a reference brick within the partial structure, thereby decomposing complex tasks into a finite set of reusable primitive skills. BrickCraft bridges the gap between high-level assembly plans and physical execution through situated manuals, which provide explicit spatial guidance for learned visuomotor skills by projecting the assembly intent onto real-time robot observations. Finally, BrickCraft employs a compositional execution pipeline that chains these spatially grounded skills to accomplish long-horizon assembly tasks. Extensive experimental validations demonstrate that BrickCraft acquires proficient assembly skills from a limited set of demonstrations and exhibits strong compositional generalization to unseen structures. The project website is available at https://intelligent-control-lab.github.io/BrickCraft.

In a conversation, a helpful assistant must reliably follow user directives, even as they refine, modify, or contradict earlier requests. Yet most instruction-following benchmarks focus on single-turn or short multi-turn scenarios, leaving open how well models handle long-horizon instruction-following tasks. To bridge this gap, we present SEQUOR, an automatic benchmark for evaluating constraint adherence in long multi-turn conversations. SEQUOR consists of simulated persona-driven interactions built with constraints extracted from real-world conversations. Our results show that even when following a single constraint, instruction-following accuracy consistently decreases as the conversation grows longer, with drops exceeding 11%. This decline becomes larger when models have to follow multiple constraints simultaneously, reducing their accuracy by over 40%. In scenarios where constraints are added or replaced at arbitrary points of the conversation, model accuracy decreases by more than 9%. Taken together, our results reveal that current models still struggle to follow user instructions in multi-turn conversations, and provide a way for better measuring instruction-following capabilities in assistants.

Coding agents are rapidly changing the landscape of software development, moving from inline completion to autonomous systems that edit repositories, open pull requests, respond to issues, and run scheduled or webhook triggered routines across the development life cycle. The next generation is increasingly described as proactive and long-horizon: agents should notice relevant changes before the developer asks, connect signals across tools, decide when to interrupt, and carry preferences across sessions. Yet the field still lacks a clear account of what proactivity means for software development, how it differs from autonomy, what acceptance criteria proactive long-horizon tasks should satisfy, and which metrics determine whether unsolicited agent behavior is useful rather than merely active. Proactive coding agents should be evaluated by the quality and improvement of their insight policy: the policy that decides what matters next, what evidence supports it, whether to show it, and how to adapt after feedback. This view is grounded in the principles of mixed initiative interaction. We propose a three level taxonomy of proactivity (Reactive, Scheduled, and Situation Aware), compare contemporary coding agents against five practical criteria, and sketch an active user simulation protocol with three evaluation targets: Insight Decision Quality (IDQ), Context Grounding Score (CGS), and Learning Lift

Recent advances in generative models have shown promise in generating behavior plans for long-horizon, sparse reward tasks. While these approaches have achieved promising results, they often lack a principled framework for hierarchical decomposition and struggle with the computational demands of real-time execution, due to their iterative denoising process. In this work, we introduce Hierarchical Diffusion-Flow (HDFlow), a novel hierarchical planning framework that optimally leverages the strengths of diffusion and rectified flow models to overcome the limitations of single-paradigm generative planners. HDFlow employs a high-level diffusion planner to generate sequences of strategic subgoals in a learned latent space, capitalizing on diffusion's powerful exploratory capabilities. These subgoals then guide a low-level rectified flow planner that generates smooth and dense trajectories, exploiting the speed and efficiency of ordinary differential equation (ODE)-based trajectory generation. We evaluate HDFlow on four challenging furniture assembly tasks in both simulation and real-world, where it significantly outperforms state-of-the-art methods. Furthermore, we also showcase our method's generalizability on two long-horizon benchmarks comprising diverse locomotion and manipulation tasks. Project website: https://hdflow-page.github.io/

Procedural tasks with multiple ordered steps are ubiquitous in daily life. Recent advances in multimodal large language models (MLLMs) have enabled personal assistants that support daily activities. However, existing systems primarily provide reactive guidance triggered by user queries, or limited proactive assistance for isolated short-term events rather than long-horizon procedural tasks. In this work, we introduce Pro$^2$Assist, a step-aware proactive assistant that continuously tracks fine-grained task progress and reasons over the user's evolving state to provide timely assistance throughout tasks. Pro$^2$Assist leverages multimodal data from augmented reality (AR) glasses to achieve motion-based perception. It then extracts step-oriented procedural context from multi-scale temporal dynamics and task-specific expert knowledge. Based on both sensory input and procedural context, Pro$^2$Assist performs continuous reasoning to infer user needs and display timely assistance on AR glasses. We evaluate Pro$^2$Assist using a dataset curated from public sources and a real-world dataset collected on our testbed with AR glasses. Extensive evaluations show that Pro$^2$Assist outperforms the best-performing baselines by over 21% in procedural action understanding accuracy, and it achieves up to 2.29x the proactive timing accuracy of baselines. A user study with 20 participants further shows that 90% find Pro$^2$Assist useful, indicating its effectiveness for real-world procedural assistance.

Modern language agents must operate over long-horizon, multi-turn histories, yet deploying such agents with Small Language Models (SLMs) remains fundamentally difficult. Full-context prompting causes context overflow, flat retrieval exposes the model to noisy evidence, and open-ended agentic loops are unreliable under limited reasoning capacity. We argue that a substantial portion of SLM memory failure arises from mismatched memory operations: different query types demand categorically different retrieval strategies, evidence transformations, and context budgets that SLMs cannot reliably self-orchestrate through open-ended reasoning. We introduce MemFlow, a training-free memory orchestration framework that externalizes memory planning from the SLM. A Router Agent classifies each query by intent and dispatches it to the Memory Agent, which executes one of three specialized tiers (Profile Lookup, Targeted Retrieval, or Deep Reasoning) and assembles the resulting evidence under a dynamic, tier-aware token budget. An Answer Agent then generates a response from this compact context, and a Validator Agent optionally retries with a heavier memory tier when the response is not supported by the provided evidence. This route-then-compile design avoids tool-selection hallucination and reasoning loops while keeping the answer context compact. Evaluated on a frozen Qwen3-1.7B backbone across long-horizon memory benchmarks - LongMemEval, LoCoMo, and LongBench - MemFlow improves accuracy by nearly 2x over full-context SLM baselines. These results suggest that structured intent routing and deterministic evidence preparation can make limited-capacity models substantially more effective in resource-constrained long-horizon agents.

This report describes ARIS (Auto-Research-in-sleep), an open-source research harness for autonomous research, including its architecture, assurance mechanisms, and early deployment experience. The performance of agent systems built on LLMs depends on both the model weights and the harness around them, which governs what information to store, retrieve, and present to the model. For long-horizon research workflows, the central failure mode is not a visible breakdown but a plausible unsupported success: a long-running agent can produce claims whose evidential support is incomplete, misreported, or silently inherited from the executor's framing. Therefore, we present ARIS as a research harness that coordinates machine-learning research workflows through cross-model adversarial collaboration as a default configuration: an executor model drives forward progress while a reviewer from a different model family is recommended to critique intermediate artifacts and request revisions. ARIS has three architectural layers. The execution layer provides more than 65 reusable Markdown-defined skills, model integrations via MCP, a persistent research wiki for iterative reuse of prior findings, and deterministic figure generation. The orchestration layer coordinates five end-to-end workflows with adjustable effort settings and configurable routing to reviewer models. The assurance layer includes a three-stage process for checking whether experimental claims are supported by evidence: integrity verification, result-to-claim mapping, and claim auditing that cross-checks manuscript statements against the claim ledger and raw evidence, as well as a five-pass scientific-editing pipeline, mathematical-proof checks, and visual inspection of the rendered PDF. A prototype self-improvement loop records research traces and proposes harness improvements that are adopted only after reviewer approval.

Large language models (LLMs) have shown promise as interactive agents that solve tasks through extended sequences of environment interactions. While prior work has primarily focused on system-level optimizations or algorithmic improvements, the role of task horizon length in shaping training dynamics remains poorly understood. In this work, we present a systematic empirical study that examines horizon length through controlled task constructions. Specifically, we construct controlled tasks in which agents face identical decision rules and reasoning structures, but differ only in the length of action sequences required for successful completion. Our results reveal that increasing horizon length alone constitutes a training bottleneck, inducing severe training instability driven by exploration difficulties and credit assignment challenges. We demonstrate that horizon reduction is a key principle to address this limitation, stabilizing training and achieving better performance in long-horizon tasks. Moreover, we find that horizon reduction is related to stronger generalization across horizon lengths: models trained under reduced horizons generalize more effectively to longer-horizon variants at inference time, a phenomenon we refer to as horizon generalization.

Many sequential decision-making problems exhibit hierarchical structure, where high-level semantic choices constrain downstream actions and feedback is delayed and ambiguous. Learning in such settings is challenging due to credit assignment: performance degradation may arise from flawed abstractions, suboptimal execution, or their interaction. We study this challenge through pair trading, a domain that naturally combines long-horizon semantic reasoning for asset pair selection with short-horizon execution under partial observability. We formulate pair trading as a hierarchical reinforcement learning problem and propose a language-driven optimization framework in which both high-level and low-level policies are parameterized by large language models (LLMs) and optimized exclusively through prompt updates. Our approach leverages pretrained LLMs as hierarchical policies and uses trajectory- and episode-level textual feedback to adapt abstractions and execution without gradient-based fine-tuning. By explicitly separating abstraction selection from execution, the framework reduces non-stationarity across hierarchical levels and enables targeted adaptation under delayed feedback. Experiments on real-world market data show consistent improvements over traditional and LLM-based baselines, demonstrating the effectiveness of language-driven hierarchical reinforcement learning.

Vision-Language-Action (VLA) models have emerged as a powerful paradigm for embodied intelligence, enabling robots to perform tasks based on natural language instructions and current visual input. However, existing VLA models struggle with long-horizon tasks due to compounding errors. Prior methods decompose tasks into subtasks of fixed granularity, which cannot adapt to the varying complexity of execution states, limiting their robustness in long-horizon tasks. To overcome this, we introduce Anticipation Model, which adaptively and recursively generates future subgoals. This model continuously adapts as the task unfolds, adjusting future subgoals in response to evolving dynamics, facilitating more reliable planning paths. Building on this concept, we propose Anticipation-VLA, a hierarchical VLA model that leverages the anticipation model to generate actionable subgoals that guide VLA policy execution. We implement Anticipation-VLA with finetuning a Unified Multimodal Model (UMM) for high-level subgoal generation and a goal-conditioned VLA policy for low-level action execution. Experiments in both simulated and real-world robotic tasks demonstrate the effectiveness of Anticipation-VLA, highlighting the importance of adaptive and recursive subgoal generation for robust policy execution.

Long-horizon LLM tasks often fail not because a single answer is unattainable, but because knowledge states drift across rounds, intermediate commitments remain implicit, and interruption fractures the evolving evidence chain. This paper presents ADEMA as a knowledge-state orchestration architecture for long-horizon knowledge synthesis rather than as a generic multi-agent runtime. The architecture combines explicit epistemic bookkeeping, heterogeneous dual-evaluator governance, adaptive task-mode switching, reputation-shaped resource allocation, checkpoint-resumable persistence, segment-level memory condensation, artifact-first assembly, and final-validity checking with safe fallback. Evidence is drawn entirely from existing materials: a four-scenario showcase package, a fixed 60-run mechanism matrix, targeted micro-ablation and artifact-chain supplements, and a repaired protocol-level benchmark in which code-oriented evaluation is the clearest quality-sensitive mechanism block. Across the fixed matrix, removing checkpoint/resume produced the only invalid run, and it did so in the interruption-sensitive resume condition. By contrast, dual evaluation, segment synthesis, and dynamic governance are best interpreted as supporting control mechanisms that shape trajectory discipline, explicit artifact progression, and cost-quality behavior rather than as universal binary prerequisites for completion. The contribution is therefore a knowledge-state orchestration architecture in which explicit epistemic state transition, evidence-bearing artifact progression, and recoverable continuity are the primary design commitments.

Existing web agent benchmarks have largely converged on short, single-site tasks that frontier models are approaching saturation on. However, real world web use consists of long-horizon, multi-site workflows. Common web navigation tasks, such as comparing products across different domains, planning trips across multiple services, or summarizing information from multiple search queries, require sustained context and cross-site reasoning over potentially hours of browsing. To capture and evaluate such behaviors, we introduce Odysseys: a benchmark of 200 long-horizon web tasks derived from real world browsing sessions evaluated on the live Internet. We find that binary pass/fail evaluation is inadequate for long-horizon settings and introduce a rubric-based evaluation, annotating each Odysseys task with an average of 6.1 graded rubrics. We demonstrate that this yields higher agreement with humans and provides a more fine-grained signal than commonly used trajectory-level LLM-as-a-judge evaluation metrics. We tested several leading frontier models and find that the strongest models achieve a success rate of 44.5%, which leaves substantial room for future improvements. Beyond task success, we argue that efficiency is a first-class concern for long-horizon agents. We introduce a Trajectory Efficiency metric (rubric score per step) and find that even frontier agents achieve only 1.15%, marking an evident need for agents that can succeed efficiently and not simply eventually. Odysseys isolates the critical evaluation of long-horizon proficiency in open-web environments, providing a realistic benchmark to measure progress towards computer-use agents that can potentially productively operate for hours. We release our tasks, evaluation scripts, and other results at https://odysseys-website.pages.dev

Cross-domain task-oriented dialogue requires reasoning over implicit and explicit feasibility constraints while planning long-horizon, multi-turn actions. Large language models (LLMs) can infer such constraints but are unreliable over long horizons, while Reinforcement learning (RL) optimizes long-horizon behavior yet cannot recover constraints from raw dialogue. Naively coupling LLMs with RL is therefore brittle: unverified or unstructured LLM outputs can corrupt state representations and misguide policy learning. Motivated by this, we propose Verified LLM-Knowledge empowered RL (VLK-RL), a hybrid framework that makes LLM-derived constraint reasoning usable for RL. VLK-RL first elicits candidate constraints with an LLM and then verifies them via a dual-role cross-examination procedure to suppress hallucinations and cross-turn inconsistencies. The verified constraints are mapped into ontology-aligned slot-value representations, yielding a structured, constraint-aware state for RL policy optimization. Experiments across multiple benchmarks demonstrate that VLK-RL significantly improves generalization and robustness, outperforming strong single-model baselines on long-horizon tasks.

Despite remarkable progress in Vision--Language--Action (VLA) models, a central bottleneck remains underexamined: the data infrastructure that underlies embodied learning. In this survey, we argue that future advances in VLA will depend less on model architecture and more on the co-design of high-fidelity data engines and structured evaluation protocols. To this end, we present a systematic, data-centric analysis of VLA research organized around three pillars: datasets, benchmarks, and data engines. For datasets, we categorize real-world and synthetic corpora along embodiment diversity, modality composition, and action space formulation, revealing a persistent fidelity-cost trade-off that fundamentally constrains large-scale collection. For benchmarks, we analyze task complexity and environment structure jointly, exposing structural gaps in compositional generalization and long-horizon reasoning evaluation that existing protocols fail to address. For data engines, we examine simulation-based, video-reconstruction, and automated task-generation paradigms, identifying their shared limitations in physical grounding and sim-to-real transfer. Synthesizing these analyses, we distill four open challenges: representation alignment, multimodal supervision, reasoning assessment, and scalable data generation. Addressing them, we argue, requires treating data infrastructure as a first-class research problem rather than a background concern.

Long-horizon manipulation remains challenging for vision-language-action (VLA) policies: real tasks are multi-step, progress-dependent, and brittle to compounding execution errors. We present LoHo-Manip, a modular framework that scales short-horizon VLA execution to long-horizon instruction following via a dedicated task-management VLM. The manager is decoupled from the executor and is invoked in a receding-horizon manner: given the current observation, it predicts a progress-aware remaining plan that combines (i) a subtask sequence with an explicit done + remaining split as lightweight language memory, and (ii) a visual trace -- a compact 2D keypoint trajectory prompt specifying where to go and what to approach next. The executor VLA is adapted to condition on the rendered trace, thereby turning long-horizon decision-making into repeated local control by following the trace. Crucially, predicting the remaining plan at each step yields an implicit closed loop: failed steps persist in subsequent outputs, and traces update accordingly, enabling automatic continuation and replanning without hand-crafted recovery logic or brittle visual-history buffers. Extensive experiments spanning embodied planning, long-horizon reasoning, trajectory prediction, and end-to-end manipulation in simulation and on a real Franka robot demonstrate strong gains in long-horizon success, robustness, and out-of-distribution generalization. Project page: https://www.liuisabella.com/LoHoManip

Vision-language-action models (VLAs) have been extensively used in robotics applications, achieving great success in various manipulation problems. More recently, VLAs have been used in long-horizon tasks and evaluated on benchmarks, such as BEHAVIOR1K (B1K), for solving complex household chores. The common metric for measuring progress in such benchmarks is success rate or partial score based on satisfaction of progress-agnostic criteria, meaning only the final states of the objects are considered, regardless of the events that lead to such states. In this paper, we argue that using such evaluation protocols say little about safety aspects of operation and can potentially exaggerate reported performance, undermining core challenges for future real-world deployment. To this end, we conduct a thorough analysis of state-of-the-art models on the B1K Challenge and evaluate policies in terms of robustness via reproducibility and consistency of performance, safety aspects of policies operations, task awareness, and key elements leading to the incompletion of tasks. We then propose evaluation protocols to capture safety violations to better measure the true performance of the policies in more complex and interactive scenarios. At the end, we discuss the limitations of the existing VLAs and motivate future research.

Long horizon interactive environments are a testbed for evaluating agents skill usage abilities. These environments demand multi step reasoning, the chaining of multiple skills over many timesteps, and robust decision making under delayed rewards and partial observability. Games are a good testbed for evaluating agent skill usage in environments. Large Language Models (LLMs) offer a promising alternative as game playing agents, but they often struggle with consistent long horizon decision making because they lack a mechanism to discover, retain, and reuse structured skills across episodes. We present COSPLAY, a co evolution framework in which an LLM decision agent retrieves skills from a learnable skill bank to guide action taking, while an agent managed skill pipeline discovers reusable skills from the agents unlabeled rollouts to form a skill bank. Our framework improves both the decision agent to learn better skill retrieval and action generation, while the skill bank agent continually extracts, refines, and updates skills together with their contracts. Experiments across six game environments show that COSPLAY with an 8B base model achieves over 25.1 percent average reward improvement against four frontier LLM baselines on single player game benchmarks while remaining competitive on multi player social reasoning games.

Industrial robotic manipulation demands reliable long-horizon execution across embodiments, tasks, and changing object distributions. While Vision-Language-Action models have demonstrated strong generalization, they remain fundamentally reactive. By optimizing the next action given the current observation without evaluating potential futures, they are brittle to the compounding failure modes of long-horizon tasks. Cortex 2.0 shifts from reactive control to plan-and-act by generating candidate future trajectories in visual latent space, scoring them for expected success and efficiency, then committing only to the highest-scoring candidate. We evaluate Cortex 2.0 on a single-arm and dual-arm manipulation platform across four tasks of increasing complexity: pick and place, item and trash sorting, screw sorting, and shoebox unpacking. Cortex 2.0 consistently outperforms state-of-the-art Vision-Language-Action baselines, achieving the best results across all tasks. The system remains reliable in unstructured environments characterized by heavy clutter, frequent occlusions, and contact-rich manipulation, where reactive policies fail. These results demonstrate that world-model-based planning can operate reliably in complex industrial environments.

Vision-Language-Action (VLA) models fail systematically on long-horizon manipulation tasks despite strong short-horizon performance. We show that this failure is not resolved by extending context length alone in the current reactive execution setting; instead, it stems from three recurring execution-loop deficiencies: the memory gap, the verification gap, and the recovery gap. We present HELM, a model-agnostic framework that addresses these deficiencies with three components: an Episodic Memory Module (EMM) that retrieves key task history via CLIP-indexed keyframes, a learned State Verifier (SV) that predicts action failure before execution from observation, action, subgoal, and memory-conditioned context, and a Harness Controller (HC) that performs rollback and replanning. The SV is the core learning contribution: it consistently outperforms rule-based feasibility checks and ensemble uncertainty baselines, and its effectiveness depends critically on access to episodic memory. On LIBERO-LONG, HELM improves task success rate by 23.1 percentage points over OpenVLA (58.4% to 81.5%), while extending the context window to H=32 yields only a 5.4-point gain and same-budget LoRA adaptation remains 12.2 points below HELM. HELM also improves long-horizon performance on CALVIN and substantially boosts recovery success under controlled perturbations. Ablations and mechanism analyses isolate the contribution of each component, and we release LIBERO-Recovery as a perturbation-injection protocol for evaluating failure recovery in long-horizon manipulation.

Image-goal navigation steers an agent to a target location specified by an image in unseen environments. Existing methods primarily handle this task by learning an end-to-end navigation policy, which compares the similarities of target and observation images and directly predicts the actions. However, when the target is distant or lies in another room, such methods fail to extract informative visual cues, leading the agent to wander around. Motivated by the human cognitive principle that deliberate, high-level reasoning guides fast, reactive execution in complex tasks, we propose Hierarchical Reasoning Navigation (HRNav), a framework that decomposes image-goal navigation into high-level planning and low-level execution. In high-level planning, a vision-language model is trained on a self-collected dataset to generate a short-horizon plan, such as whether the agent should walk through the door or down the hallway. This downgrades the difficulty of the long-horizon task, making it more amenable to the execution part. In low-level execution, an online reinforcement learning policy is utilized to decide actions conditioned on the short-horizon plan. We also devise a novel Wandering Suppression Penalty (WSP) to further reduce the wandering problem. Together, these components form a hierarchical framework for Image-Goal Navigation. Extensive experiments in both simulation and real-world environments demonstrate the superiority of our method.

Long-horizon large language model (LLM) agents are fundamentally limited by context. As interactions become longer, tool descriptions, retrieved memories, and raw environmental feedback accumulate and push out the information needed for decision-making. At the same time, useful experience gained from tasks is often lost across episodes. We argue that long-horizon performance is determined not by context length, but by how much decision-relevant information is maintained within a finite context budget. We present GenericAgent (GA), a general-purpose, self-evolving LLM agent system built around a single principle: context information density maximization. GA implements this through four closely connected components: a minimal atomic tool set that keeps the interface simple, a hierarchical on-demand memory that only shows a small high-level view by default, a self-evolution mechanism that turns verified past trajectories into reusable SOPs and executable code, and a context truncation and compression layer that maintains information density during long executions. Across task completion, tool use efficiency, memory effectiveness, self-evolution, and web browsing, GA consistently outperforms leading agent systems while using significantly fewer tokens and interactions, and it continues to evolve over time. Project: https://github.com/lsdefine/GenericAgent

Robotic manipulation policies often degrade over extended horizons, yet existing benchmarks provide limited insight into why such failures occur. Most prior benchmarks are either simulation-based or report aggregate success, making it difficult to disentangle the distinct sources of temporal difficulty in real-world execution. We introduce LongBench, a real-world benchmark for evaluating long-horizon manipulation. LongBench consists of over 1,000 real-world episodes, covering two complementary regimes: Context-Independent (fully observable) and Context-Dependent (ambiguity-driven). By organizing tasks into capability- and ambiguity-specific subsets, LongBench enables mechanism-aware evaluation of execution robustness, temporal consistency, and context-dependent reasoning. Evaluating six state-of-the-art policies reveals that long-horizon performance is not governed by a single factor. We observe that performance in fully observable settings is more strongly associated with execution robustness, while contextual difficulty varies across tasks and is not consistently improved by memory-based methods. We hope that LongBench serves as a useful benchmark for studying long-horizon manipulation and for developing policies with stronger robustness across both execution and contextual challenges.