每天知道
你的领域
出了什么

Generating novel renderings of a scene along user-defined camera trajectories from a single monocular video, dubbed video retaking, is a compelling but difficult problem in content creation and visual effects. Existing geometry-guided approaches reconstruct a 4D representation from the source video and render it along the target trajectory to condition video diffusion models. However, this guidance degrades as the target camera departs from the source trajectory, leaving newly revealed regions sparse or entirely missing. We propose SierpinskiCam, which addresses this limitation by augmenting geometry-based guidance with Sierpinski dome texture cues that contains rich trackable features even under large viewpoint changes. We further introduce a reference video conditioning mechanism that appends source-video tokens to the target-token sequence and separates the two streams with negative RoPE indices, enabling appearance grounding without architectural modification or per-video adaptation. Extensive experiments show that SierpinskiCam achieves significant gains in camera controllability, geometric consistency, and video quality across diverse and challenging retaking scenarios. Project page: https://hyelinnam.github.io/SierpinskiCam/.

World models are transitioning from passive visual generators to foundational, operational infrastructure for Physical AI: they must natively acquire world knowledge from heterogeneous experience, maintain persistent states over long horizons, and execute efficiently within real deployment constraints. We introduce Kairos, a native world model stack designed around these requirements. (1) Kairos learns the world by pioneering a Native Pre-training Paradigm governed by a Cross-Embodiment Data Curriculum, which organizes open-world videos, human behavioral data, and robot interactions into a progressive developmental pathway. (2) Kairos maintains the world by unified world understanding, generation, and prediction within a Native Unified Architecture equipped with Hybrid Linear Temporal Attention, where sliding-window attention captures local dynamics, dilated sliding windows capture mid-range dependencies, and gated linear attention maintains persistent global memory. We establish formal theoretical bounds demonstrating that this temporal factorization strictly limits error accumulation, mathematically guaranteeing state propagation across extended horizons. (3) Kairos runs the world by incorporating a Deployment-Aware System Co-Design to support low-latency rollout generation on server and consumer-grade hardware for real-world observation-action-feedback loops. Experiments on embodied world-model, long-horizon, and action-policy benchmarks show that Kairos achieves top level performance while offering a strong efficiency-capability trade-off. Together, these results position Kairos as a cohesive operational foundation for future self-evolving physical intelligence.

We introduce Qwen-RobotWorld, a language-conditioned video world model for embodied intelligence. With natural language as a unified action interface, it predicts physically grounded future visual trajectories from current observations across robotic manipulation, autonomous driving, indoor navigation, and human-to-robot transfer. This unified formulation provides three promising application directions: synthetic data generation for policy training augmentation, scalable virtual environments for policy evaluation, and language-guided planning signals for downstream robot control. This is achieved through a three-part design: a) Double-Stream MMDiT with MLLM Action Encoding, where a 60-layer double-stream diffusion transformer couples frozen Qwen2.5-VL semantics with video-VAE latents through layer-wise joint attention; b) Embodied World Knowledge (EWK), an 8.6M video-text corpus (200M+ frames) with action-language mapping over 20+ embodiments and 500+ action categories; and c) General+Expert Progressive Curriculum, a two-stage training strategy that first learns general visual priors and then injects embodied specialization under a shared language interface. Extensive results show strong competitiveness: ranks 1st overall on EWMBench and DreamGen Bench, outperforms all open-source models on WorldModelBench and PBench. Additional zero-shot analyses on RoboTwin-IF benchmark further support robust generalization and multi-view consistency.

AI systems based on artificial neural networks are being developed with aspirations of pushing the boundary of human mathematical knowledge. A key question for these systems is how much they can reach beyond their training data. Mathematical discovery requires a strong form of out of distribution generalization; the ability to hypothesize genuinely new - and potentially logically more powerful - mathematical structures. It has been hypothesized that language abilities support such generalizations in human cognition. In this work, we use simple arithmetic as a case study for examining how modern AI models could expand their mathematical horizons, evaluating whether these models can independently discover the concept of "zero". We show that We show that (1) language models of a GPT-2 size are unable to perform this generalization at test time regardless of language pretraining, but (2) models can improve substantially after training on tens or hundreds of examples of zero. Additionally, we find that language pretraining reduces the number of required examples by approximately $50%$, showing that language abilities can scaffold mathematical discovery in neural models.

Organizations and researchers show increasing interest in using large language models (LLMs) in place of human participants in A/B tests, in the hope of experimenting faster and at lower cost. We study when a treatment effect estimated on LLM outcomes recovers the effect that would have been measured on the human population of interest. Distributional equivalence between LLM and human outcomes would make any standard estimator valid but is unrealistic. We therefore develop a statistical framework that adapts surrogate endpoint theory to LLMs. The framework shows that calibrating LLM outcomes to human outcomes identifies the average treatment effect under surrogacy and comparability conditions that are jointly weaker than distributional equivalence. When these conditions fail, the effect of interest is only partially identified, and we provide diagnostics that can falsify surrogacy on historical experiments together with a bound on the worst-case bias from limited overlap. We further show that the stochasticity inherent to LLMs introduces both bias and variance, but using an average of multiple draws as the surrogate mitigates both. We illustrate the methods and theory in simulations and an application to A/B tests on Upworthy headlines. A central takeaway from our work is that the validity of LLM outcomes as surrogates can only be falsified for past treatments and never verified for new ones, so human experiments remain indispensable for novel interventions. We discuss the role of LLM choice, prompting, and temperature as design variables, and how to size human experiments for validation.

Pretraining language models with extended context windows enhances their ability to leverage rich information during generation. Existing methods split input sequences into chunks, broadcast them across multiple devices, and compute attention block by block which incurs significant communication overhead. While feasible in high-speed clusters, these methods are impractical for decentralized training over low-bandwidth connections. We propose a compression method for communication-efficient context parallelism in decentralized settings, achieving a remarkable compression rate of over 95% with negligible overhead and no loss in convergence. Our key insight is to exploit the intrinsic low-rank structure of activation outputs by dynamically constraining them to learned mixtures of subspaces via efficient reparameterizations. We demonstrate scaling billion-parameter decentralized models to context lengths exceeding 100K tokens on networks as slow as 300Mbps, matching the wall-clock convergence speed of centralized models on 100Gbps interconnects.

LLM agents mis-call tools, and the natural guess is that the model failed to see the right tool in a crowded harness. We show the opposite through a lens concurrent work sets aside — the model's attention to labeled tool-definition segments. On real BFCL failures, by per-candidate attention argmax the model attends most to the correct tool 80% of the time (vs. 21% chance), and the gold is the under-attended segment on only 10%: it looks at the right tool and still picks wrong. This directly refutes the intuitive "crowded-harness / lost-in-the-middle" explanation: the failure is at the decision readout, not the harness, and we pin it there three ways. (1) Input vs. readout: repairing the prompt (reordering or duplicating the gold tool) recovers <=23% of failures, while readout-side interventions recover 59-91%. (2) Representation-invariance: two gold-pointed interventions in different representations — an additive attention-logit bias and a residual-stream steering vector — recover largely the same failures (per-task Jaccard 0.865 pooled, 0.79-0.91 per model), so the bottleneck is localized to the readout independent of which representation is poked. (3) A training-free, gold-free selector: per-segment attention closes most of the gold-free-vs-oracle gap on BFCL (+11.9 pts pooled function-name selection vs. +17.9-pt oracle headroom) and adds +14.9 pts on Seal-Tools; every model positive (exact McNemar p<=8e-4 each). Scopes differ: the causal attention-bias dose-response is bidirectional and monotonic on 10 mask-honoring models (3-32B), the full 0.5-32B span carrying only the correlational diagnostic; the deployable selector is evaluated on 5 single-turn models and does not yet transfer to a multi-turn loop.

LLM-based agents have demonstrated strong capabilities in data-intensive analytical tasks, yet their outputs are rarely verifiable: a reliance on linear text trajectories makes their reasoning difficult to audit. In particular, deterministic computations over raw data and semantic deductions over natural-language claims are often entangled in an unstructured stream, leaving numerical conclusions hard to reproduce and qualitative judgments hard to inspect. To address this, we propose VeriGraph, a traceable neuro-symbolic reasoning framework that enables agents to construct an explicit heterogeneous evidence directed acyclic graph (DAG) during execution. VeriGraph introduces three evidence-expansion primitives, namely computational, grounding, and derivational expansion, to connect raw data, interpreter variables, computed results, and natural-language claims in a unified graph. Under this formulation, structural traceability is reduced to graph reachability from raw data sources to terminal claims, while semantic support is measured by claim-level evidence evaluation. To improve graph construction, we further design a graph-based policy optimization strategy with a composite reward that jointly supervises answer correctness, computational integrity, and derivational coherence. Experiments on four benchmarks show that VeriGraph-8B achieves the highest overall score among all baselines. More importantly, VeriGraph produces auditable evidence graphs with substantially stronger claim grounding, achieving a 87.61% Grounding Rate under our claim-level evidence support evaluation. These results suggest that explicit evidence-graph construction is a promising path toward verifiable data-analytic agents. Our code is available at https://github.com/ignorejjj/VeriGraph.

When AI agents use language models to evaluate their own outputs in a feedback loop, systematic biases emerge. We show that Evaluator Preference Collapse (EPC) is dramatically amplified in multimodal settings. Using GPT-4o to evaluate DeepSeek-chat across text and visual tasks, we find that a single strategy (step_by_step) absorbs 48.4% of all weight — 3.2x the collapse observed in text-only self-evaluation — while three visual-domain strategies receive only 9.1% combined weight. We then demonstrate a novel phenomenon we term cross-modal contagion: evaluator preferences acquired on one modality transfer to and corrupt strategy selection on another. Through a four-phase isolation training paradigm, we measure contagion coefficients and document strategy inversion — the optimal strategy for a modality reverses after cross-modal exposure. A Phase 3 statistical validation across four evaluator configurations (N=53 total independent repetitions, 15,592 API calls) reveals a clear hierarchy: cross-model evaluation (GPT-4o, N=8) produces strong but symmetric bidirectional contagion (mean gamma_{T->V}=1.176, gamma_{V->T}=1.089, Delta=-0.088, p=0.575, Cohen's d=0.29); high round counts (DashScope, 50 rounds) cause collapse to single-strategy dominance (70% zero contagion); and self-evaluation provides near-complete immunity — 97% of runs (N=30, DeepSeek-chat) yield exactly zero contagion (mean gamma=0.033, 95% CI [-0.031, 0.010], p=0.642, d=0.07). No evaluator condition shows statistically significant directional asymmetry. We introduce the contagion matrix indexed by evaluator identity, release the MM-EPC experimental framework, and identify cross-model evaluator architecture as the primary risk factor for preference contagion.

Deep research agents synthesize long-form reports by searching and reasoning over retrieved evidence. Reinforcement learning with rubric-based rewards improves these agents by optimizing them against checkable criteria that translate report quality into reward signals, but its efficiency depends on whether those criteria reliably capture the task scope and evidence needs. Most existing studies ask an LLM to generate rubrics for a given query, but when the model fails to infer the underlying information needs, the generated rubrics may be incomplete and reduce RL efficiency. To obtain more reliable query--rubric supervision, we introduce DeepRubric, a data construction framework that reverses this process: instead of inferring evaluation criteria for a given query, it first determines what an evidence-backed report should be evaluated on and then synthesizes aligned query--rubric pairs from those evaluation targets. Starting from a sampled seed topic, DeepRubric builds an evidence tree by recursively expanding evidence-backed sub-questions, whose leaves serve as atomic and verifiable evaluation targets. It then uses the evidence tree to synthesize the training query and rubrics, ensuring that the reward evaluates exactly the information requested by the query. Using DeepRubric, we construct 9K query--rubric supervision examples and train DeepRubric-8B with rubric-based GRPO, achieving comparable performance to prior open state-of-the-art deep research models across three benchmarks with roughly 13x fewer RL GPU-hours.

Long-horizon agents rely on memory mechanisms to compress interaction history, but optimizing memory writing faces a distinct credit assignment challenge: a memory update may be rewarded or penalized due to downstream tool failures, noisy observations, or reasoning errors rather than its own contribution. This causally entangled credit can lead agents to discard useful evidence or preserve irrelevant information. We propose HiMPO, a Hindsight-Informed Memory Policy Optimization framework for assigning less-entangled credit to memory-writing actions in long-horizon agents. HiMPO first estimates the local utility of a memory update by comparing the task-relevant information recoverable from the previous and updated memories under the same pre-write state. It then uses hindsight relevance as a bounded retrospective filter that attenuates memory credit when local utility is not supported by the target outcome. The resulting memory-specific advantage is applied only to memory tokens, while trajectory-level rewards optimize the rest of the agent behavior. Across judge-based open-domain tasks and objective compressive-memory QA, HiMPO improves over strong memory-based and RL-based baselines while preserving compressed-context efficiency. Controlled interventions further show that HiMPO reduces blame leakage from tool-induced errors and improves attribution fidelity of memory updates.

Tool-augmented LLM agents rely on runtime filtering to decide which tools should be visible at each step. Causal Minimal Tool Filtering (CMTF) reduces tool-choice confusion by exposing only the next causally necessary tool frontier, but it assumes that the user request has already been mapped to a symbolic goal state. In practice, requests such as "handle my appointment" or "take care of this email" may correspond to multiple possible goals. This creates wrong-goal execution, where an agent follows a valid causal tool path for an unintended objective. We introduce GIST-CMTF, a goal-state inference layer that predicts candidate symbolic goals over the same state-transition vocabulary used by CMTF, estimates ambiguity, and either applies CMTF or exposes clarification as a causal action that produces missing goal or state variables. We evaluate GIST-CMTF across seven model backends, six filtering methods, and 120 controlled tool-use tasks. GIST-CMTF achieves 97.0% task success, compared with 80.1% for top-goal CMTF and 82.9% for semantic-goal CMTF. It reduces wrong-goal execution from 19.4% under top-goal CMTF to 2.5%, while preserving the one-tool exposure of causal filtering and using substantially fewer tokens than all-tools exposure. These results suggest that reliable tool-augmented agents should validate goal state, not only tool relevance, before exposing external actions.

Training tool-augmented LLM agents requires large corpora of multi-turn, tool-grounded conversational data that is expensive to annotate, privacy-constrained in production settings, and largely absent from public datasets. We present StateGen, a synthetic data generation platform that produces scored, reasoning-trace-rich training conversations by orchestrating a four-role LLM loop: a persona-conditioned user simulator, an agent under test, a state-grounded tool simulator, and a multi-axis LLM judge. The key architectural contribution is an authoritative state manager that maintains a structured world-state object across turns, enforcing a backend-is-truth invariant that eliminates the dominant class of tool-call hallucinations by construction. StateGen extends naturally to hierarchical multi-agent settings by declaring sub-agents as tools, all sharing a single state object. We report results on 64,698 evaluated conversations across three production corpora: tool-call hallucination scores reach 9.66/10, the system supports persona-driven variation via a 23-dimensional trait vector, and a cleanly separated train and golden evaluation set split confirms the data is not memorization bait (per-criterion gap analysis). Comparison with eight external systems shows that no single publicly available platform combines multi-turn generation, state-grounded tool simulation, hierarchical multi-agent support, and built-in judge scoring.

User instructions are often underspecified because humans rely on implicit assumptions about the surrounding environment. For large language model (LLM) agents operating in information-rich digital and physical environments, these assumptions cannot be inferred from the instruction alone; they must be recovered from the current state of tools, data, interfaces, and observations. Effective execution therefore requires agents to identify missing context, ground it in observed evidence, and carry it forward into subsequent actions. We show that current agents often fail to do so. They act from assumed rather than observed specifics, overlook information they could have gathered, and fail to incorporate evidence that has already been returned. Building on this insight, we propose ACCORD (Action-Conditioned Contextual Grounding), a simple and effective agent framework for adaptive grounding. Before each action, ACCORD actively probes the environment for missing information and integrates relevant context from the agent's trajectory that would otherwise be overlooked. Requiring no additional training or task-success signals, ACCORD improves task-goal completion on AppWorld by up to +20.6 points with GPT-5-mini, from 42.0% to 62.6%, compared to strong baselines. These gains persist with a substantially stronger base model (+10.8 with Claude-4.5-sonnet), an open-weight model (+10.1 with Qwen3.5-27B-FP8), and on the embodied AlfWorld benchmark (+7.4 success rate with GPT-5-mini).

AI agents can now take irreversible actions in operational systems, but agent-caused losses are still not clearly assigned, priced, or transferred. Providers often disclaim consequential damages, users are left with uncompensated losses, and default human review limits the efficiency gains of automation. We ask when autonomous AI deployment can become economically acceptable despite failure risk. Our answer is to quantify risk at the customer-task-trace episode level and transfer it through insurance. Automation is acceptable when its expected benefit exceeds the premium, control cost, and remaining risk. This requires a defined role with bounded permissions and comparable traces. We introduce trace-economic underwriting, which maps tool-use traces to customer exposure and claimable loss, then uses this representation for pricing, control, and risk transfer. It uses deterministic economic labels rather than an LLM judge. In our trace-to-loss testbed, trace-economic pricing reduces pricing MAE from $17.7K to $569 and removes regressive cross-subsidy. A 300-trace expert audit accepts 295 labels unchanged. On 1,000 real SWE-smith traces, trace-conditioned controls reduce CVaR95 by 72%. Theorem~1 gives a finite-sample scope condition. We release code, labels, and audit sheets.

A personalized AI agent needs a user memory: a persistent model of who the user is, built across many conversations and consulted on each new one. Today this memory is almost always stored as unstructured text, a knowledge graph, or a flat store of facts, and consulted by retrieval — fetching the entries most similar to the current request. Such "bag-of-facts" memory recalls individual facts well, but because storing a fact and acting on it are separate steps, it struggles to resolve contradictions, aggregate over many records, or enforce rules. We argue that user memory should instead be executable. We introduce User as Code (UaC), a paradigm in which an agent's model of a user is a living software project: typed Python objects hold the user's state and ordinary Python functions encode the rules that govern it, so representing and reasoning about the user happen in one medium an interpreter can run. The enabling mechanism is a two-phase pipeline: an append-only log that never discards a fact, periodically checkpointed into typed code. This changes what memory can do. On standard long-term conversation benchmarks, UaC matches both a full-context upper bound and the strongest prior memory systems on recall (78.8% on LOCOMO). Its advantage emerges where representation matters most. On aggregate questions over a user's history — "how many international trips did I take last year?" — retrieval-based memory collapses (6-43%) while UaC stays near-perfect (99%), because the answer is a one-line computation over typed state rather than a search over text. And because its rules execute deterministically whenever the state changes, UaC can surface unsolicited, safety-critical alerts — such as a newly prescribed drug that conflicts with an allergy recorded months earlier — a capability query-driven memory cannot provide.

Recent advances in tool-integrated language agents have significantly improved their ability to solve complex reasoning tasks. However, existing alignment methods predominantly focus on maximizing task accuracy, while overlooking auxiliary objectives such as tool-use efficiency, which are essential for practical deployment. To address this gap, we introduce ParetoPO, a two-stage multi-objective optimization framework for aligning tool-using large language models (LLMs) under competing objectives. In the first stage, ParetoPO leverages hypervolume-guided dynamic scalarization to adapt reward weights based on global Pareto frontier progress. In the second stage, it replaces scalarized learning signals with Pareto-ranking-based advantage computation, promoting nondominated trajectories through dominance-aware credit assignment. This design enables fine-grained, action-level optimization across multiple conflicting objectives. Experimental results on mathematic reasoning and multi-hop QA tasks show that ParetoPO consistently discovers policies with superior accuracy-efficiency trade-offs compared to static and heuristic baselines.

Effective personalized AI-assisted learning demands systems that can not only generate accurate learner-specific educational materials, but also dynamically adapt their instruction to diverse learners. However, existing educational agents have primarily focused on lecture content automation and simulations, which often fall short of modelling multimodal and embodied instructional methods tailored for the individual learner. To this end, we propose LectūraAgents - a multi-agent framework that enables personalized learning through end-to-end adaptive embodied teaching. At its core, LectūraAgents mirrors a professor-student relationship, in which a ProfessorAgent leads a collaborative team of specialized subordinate agents through research, planning, review, and embodied delivery of lecture contents that adapt to a learner's needs. The framework offers three main contributions: (1) a hierarchical multi-agent architecture for end-to-end personalized learning; (2) an adaptive embodied teaching mechanism, wherein the ProfessorAgent executes visible and pedagogically motivated teaching actions (e.g., handwrite, highlight, underline, etc.) over contents in a teaching environment; and (3) a Teaching Action-Speech Alignment (TASA) algorithm that employs salience-based heuristics and temporal semantic segmentation to generate coherent teaching action sequences aligned with learner profiles. We evaluate LectūraAgents on diverse courses at high school, undergraduate, and graduate levels using sample-specific rubric-based analysis; with generated lecture materials and teaching actions assessed and validated by expert educators. Experimental results show consistent gains in lecture content quality, embodied teaching quality, assessment, and personalization over existing approaches, positioning LectūraAgents as a pedagogically well-grounded framework for personalized learning at scale.

Large language model (LLM) agents increasingly take actions (screening applicants, recommending credit, triaging patients), yet fairness for LLMs is still measured by grading answers. We introduce AgentFairBench, a cheap, reproducible, multi-domain benchmark for demographic disparity in the actions of LLM agents. Grounded in a companion framework, the Bias Conduction Framework (BCF, restated here), it spans three regulator-anchored domains: hiring, lending, and medical triage. Synthetic, demographic-neutral profiles are evaluated in counterfactual matched sets that vary only a name-coded race x gender signal (in the Bertrand Mullainathan tradition), under four agent scaffolds of increasing agency (direct, chain-of-thought, multi-agent deliberation, tool-augmented). A NumPy-only harness computes counterfactual flip rate, mean absolute score difference (MASD), action-rate disparity, and tool-invocation disparity, with bootstrap confidence intervals, paired tests, and false-discovery-rate control, for single-digit dollars per model. A live leaderboard with a held-out private split and a contamination canary admits external models by submission. Our pilot (864 decisions plus a test-retest replication) carries a methodological lesson: comparing a six-group score spread against a two-run noise difference overstates disparity by ~ 2.4X through statistic arity alone. Against an arity matched noise floor and an omnibus group test, claude haiku 4 5 shows no demographic effect above sampling noise (0 of 120 pairwise and 0 of 9 omnibus contrasts survive correction); a planted-bias test confirms the instrument detects disparity when present. The contribution is a sound, sensitive, adoption-ready instrument, the arity matched null methodology, and open artifacts to scale it. Code, data, and harness are released under open licenses, with an anonymized review artifact.

Large language model (LLM)-based search agents synthesize open-web content into actionable recommendations on behalf of users, creating a risk that attacker-published pages are transformed into endorsed claims. We introduce SearchGEO, a controlled evaluation framework for measuring endorsement corruption in LLM-based web-search agents, combining a web-evidence manipulation pipeline, a five-mode attack taxonomy, and multiple output-level metrics. We evaluate 13 LLM backends on 308 cases each. Results show that vulnerability patterns vary across backends: overall attack success rate (ASR) ranges from 0.0% on Claude-Sonnet-4.6 to 31.4% on Gemini-3-Flash, the strongest attack mode differs by model family, and the same deployment scaffold could amplify or decrease ASR on different backends. An auxiliary agent-skill probe, where endorsement becomes an install command, exposes a sharp split among otherwise robust backends: Claude over-rejects while GPT over-trusts. These findings argue for treating recommendation reliability under adversarial search content as a first-class dimension of backend safety evaluation.

Most medical AI systems improve by scaling additional machinery: more fine-tuning data, more agents, and/or larger retrieval databases. In rare-disease diagnosis, however, such scaling can produce systems that are difficult to deploy, audit, and maintain. We asked whether state-of-the-art diagnostic performance could instead be achieved by extending the reasoning chain of a single AI agent: guiding it with a diagnostic policy, developed through human-AI collaboration and augmenting with freely available biomedical tools. We introduce LiteOdyssey, a lightweight rare-disease diagnostic framework that guides reasoning language model through a clinical genetics workflow. This framework was developed through Policy Iteration with Human Feedback (PIHF) and uses dynamic access to public biomedical tools. On two challenging benchmarks that provide only patient clinical features, LiteOdyssey achieved state-of-the-art performance, with an overall disease Recall@1 of 59.3% over the combined 1,243 cases of LIRICAL (n = 370) and the PhenoPacket Store (n = 873). Both benchmarks have a high proportion of ultra-rare disease (a prevalence below 1 in 1,000,000, with ultra-rare shares of approximately 45% and 52.8%, respectively). On the more difficult PhenoPacket subset, where causal diseases were not mapped to Orphanet in our rarity-mapping pipeline, LiteOdyssey achieved 60.7% Recall@1, compared with 10.7% for the same baseline model (GPT-5.4) without tools. This performance was achieved without fine-tuning, multi-agent ensembles, or a large case-retrieval database. Gains were also observed in the following: on cases never seen during development, on a private cohort of real-world rare disease patients, and on a smaller open-weights model. LiteOdyssey suggests a path toward rare-disease AI systems that are accurate, easier to deploy, and more transparent for physician review.

Multi-turn tool-use agents must reason, call tools, and adapt to observations across several interaction turns. Post-training such agents is challenging, as reinforcement learning often suffers from sparse rewards and weak credit assignment despite matching the prompt-only inference setting, while supervised fine-tuning on expert traces provides dense process supervision but can over-constrain the model to fixed trajectories. To tackle this, we propose PACT, a Privileged trAce Co-Training framework for multi-turn tool-use agents. The key idea is to use expert traces only as training-time optimization signals rather than rollout-time hints. PACT keeps rollout generation prompt-only, then uses expert traces to guide optimization through two complementary signals: a trace-conditioned RL surrogate that evaluates prompt-only rollouts under expert-trace context, and a component-aware SFT loss that supervises reasoning prefixes and tool-calls with annealed strength. To reduce over-reliance on the training-only trace context, PACT further introduces a prompt-only anchoring. We also provide a latent-trace view that connects the two trace-based objectives and explains how expert traces can guide optimization without being used during rollout generation. Experiments on FTRL, BFCL, and ToolHop show that PACT consistently improves over strong SFT- and RL-based baselines, highlighting the value of privileged trace co-training for multi-turn tool-use learning.

As LLM agents become capable of increasingly long-horizon tasks, evaluating their performance in economic systems is becoming increasingly important. Unlike existing benchmarks that primarily evaluate a single agent interacting with a passive environment, economic systems are inherently multi-agent, requiring autonomous agents to communicate, negotiate, and transact while pursuing their own objectives over extended periods. We introduce CoffeeBench, a benchmark for evaluating LLM agents in a long-horizon multi-agent economy composed of heterogeneous firms. In CoffeeBench, two farmers, two roasters, and two retailers autonomously operate their businesses over a 90-day simulation, each seeking to maximize cumulative net income through communication and transactions while managing cash, inventory, and pricing. The evaluated model controls one coffee roaster, while the remaining firms are controlled by fixed reference agents. Across several recent open-weight and proprietary LLMs, all models outperform a passive baseline that takes no actions, with most achieving positive net income. Analysis of agent behavior reveals substantial differences in long-horizon economic interaction: higher-performing models communicate more actively with other firms, whereas Claude~Haiku~4.5 exhibits an idle-drift failure mode, repeatedly choosing inaction despite producing coherent assessments and plans. We release our code and agent trajectories to support future research.

Art therapy plays a vital role in emotional healing, in which narrative creation acts as the primary vehicle for emotional expression. Given the inherently dynamic nature of emotions during healing, narratives with finely controlled emotional fluctuations enable individuals to safely project inner conflicts and achieve emotional catharsis. Recently, with the rapid development of Large Language Models (LLMs), automated narrative generation technology has provided a new pathway to support such artistic designs. However, while existing methods can produce fluent texts, they struggle to generate narratives that adhere to specified affective trajectories, failing to meet the demands of emotion-oriented psychological healing. To address these issues, this paper proposes EC-Script, an LLM agent-based framework that enables hierarchical control of the affective trajectory in narrative generation for emotional healing. To ensure that the generated narratives strictly follow the given emotional patterns, EC-Script establishes overall narrative direction through Emotion-Trajectory Planning, propels scene-level plot development with Character-Driven Scene Generation, and regulates local emotional changes of characters via Emotion-Controlled Script Writing. Ultimately, it outputs scene-by-scene script content that remains highly consistent with the preset affective trajectory. Experimental results demonstrate that EC-Script significantly outperforms baseline methods in affective trajectory adherence, exhibiting excellent and reliable emotional controllability, thereby providing effective technical support for AI-assisted emotional healing scenarios.

Large language models (LLMs) are increasingly being used for automated decision-making systems in finance, healthcare, or environmental monitoring. Time series data are ubiquitous in these fields, yet hard to process automatically. Can time series be analyzed by LLM agents? We examine three approaches: providing the agent with raw numerical data, using the LLM as a coding agent, or a combination of both. In the coding agent setup, the model iteratively queries the data using Python code. Using two time series understanding benchmarks, we show that agents with code access can outperform models processing raw data by up to 10%. However, even the best performing agent still answers about 22-34% of the questions incorrectly. To get insights into models' strategies and reasoning gaps, we analyze the model outputs with a strong LLM judge. Our analysis reveals that coding agents can select appropriate statistical tests, but often miss important nuances. Meanwhile, models with access to raw data can reach the right conclusions using back-of-the-envelope calculations.

Code-as-Policies (CaP) has shown that large language models (LLMs) can write code to solve robotics tasks by composing perception, planning, and control primitives. Recent CaP systems, however, rely on multi-turn code-generation loops at test time, which is often infeasible for real-time robot control. We introduce Robotics Harness Optimization (RHO), a novel paradigm in which tool-enabled coding agents, at training time, propose and search for interpretable, neurosymbolic multi-file policy repositories (Repositories-as-Policies) that compose these primitives rather than a single prompt, function, or file. RHO searches with reflective feedback from environment reward and execution rather than teleoperation demonstrations. It generalizes to perturbed pick-and-place settings like LIBERO-PRO, where OpenVLA scores 0.0% and $π_{0.5}$ averages 12.83%. Using the same low-level primitives, RHO reaches a 45.0% success rate, 2.5x higher than the strongest multi-turn agentic system, and 3.5x higher than $π_{0.5}$. On Robosuite, RHO sets a new state-of-the-art of 70.0%, exceeding the prior multi-turn record of 68.29% using single-turn execution with no corrective LLM code edits at deployment. When an LLM is used in the control loop, as on RAI's O3DE benchmark, RHO optimizes the deployed agent's multi-file harness of prompts, tools, and control code, improving held-out success from 23.5% to 44.3% with 20% less wall-clock time and 27% fewer tool calls.

Current computer-use benchmarks primarily focus on software operation tasks in virtualized systems, whereas scientific instrumentation scenarios require coordinated control over complex interfaces, and feedback-driven parameter adjustment. However, directly evaluating agents on physical high-precision instruments is impractical due to high cost, safety risks, limited accessibility, and difficulty in ensuring reproducible evaluation. This motivates the need for a simulated yet realistic testbed that preserves the operational challenges of scientific instruments while enabling scalable and safe benchmarking. To this end, we introduce LabOSBench, a challenging benchmark for multimodal GUI agents built on a suite of web-based scientific-instrument simulators. Operating directly via a browser, LabOSBench avoids resource-heavy OS virtualization while supporting flexible task configuration and execution-based evaluation. Specifically, LabOSBench constructs 96 subtasks across eight instrument simulators, covering workflows from sample loading, alignment, parameter tuning, and data acquisition to result inspection. We evaluate general-purpose vision-language models, specialized GUI agent models, and advanced agentic frameworks at both subtask and end-to-end levels. Our experiments reveal that while existing agents can complete many structured GUI subtasks, they still struggle with feedback-driven operations and long-horizon workflow execution. Overall, LabOSBench provides a reproducible, low-cost testbed for advancing computer-using agents toward scientific-instrument control.

Diffusion and flow-based generative policies provide a powerful policy class for reinforcement learning by inducing rich stochastic exploration through iterative action generation. However, the stochasticity of diffusion policies is not suitable for stable and precise control in high-dimensional robotic systems, where small action variations can accumulate into inconsistent motion and reduced robustness. To address this issue, we propose SteerGenPO, a latent-space reinforcement learning framework that steers a trained generative policy into a robust deterministic robotic controller. The key idea is to replace stochastic latent sampling of the trained generative policy with a learned latent actor that predicts a state-dependent latent input for the generative policies. This separates exploration and control: stochastic generative sampling provides diverse action proposals during policy learning, while deterministic latent steering provides stable and adaptive control at deployment. We evaluate SteerGenPO on six Isaac Lab benchmarks and a Unitree G1 locomotion task. The results show SteerGenPO improves over both classical RL and generative RL baselines, while its deterministic latent steering produces more stable inference-time behaviors and more reliable command responses.

Robots deployed in the real world must plan motions across diverse scenarios without per-scenario retuning. End-to-end reinforcement learning (RL) can generalize across scenarios but often becomes brittle under distribution shift, reward misspecification, and stochastic interactions. Model predictive path integral (MPPI) control enables strong real-time refinement without gradients, but its performance depends on a well-shaped sampling prior, while manually designing the priors does not scale to multi-scenario deployment. We present HOLO-MPPI (High-level Offline, Low-level Online MPPI), a multi-scenario motion planning framework that combines high-level policy learning with low-level stochastic optimal control. Offline, we learn a high-level policy that proposes scenario-robust plans in an abstract action space, with a learned world model for online rollout. Online, the policy serves as a data-driven prior generator that parameterizes MPPI's sampling distribution conditioned on the current observation and goal. MPPI then optimizes low-level control sequences around this prior in real time to adapt to local disturbances. We instantiate HOLO-MPPI in autonomous driving by designing an effective high-level action space and tailored model architectures. Our evaluation across diverse driving scenarios shows that HOLO-MPPI improves upon MPPI and end-to-end RL baselines while maintaining real-time control.

Conveying complex objectives to reinforcement learning (RL) agents often requires meticulous reward engineering. Preference-based RL (PbRL) offers a promising alternative by learning reward functions from human feedback, but its scalability is hindered by high labeling costs. Inspired by advances in Video Foundation Models (ViFMs), we present Video-based Optimal Transport Preference (VOTP), a semi-supervised framework that learns effective reward functions from only a handful of labels. By leveraging optimal transport to align visual trajectories within the rich representation space of ViFMs, VOTP effectively generates high-fidelity pseudo-labels for large amounts of unlabeled data, substantially reducing human supervision. Extensive experiments across locomotion and manipulation benchmarks demonstrate the superiority of VOTP, which outperforms state-of-the-art offline PbRL methods under limited feedback budgets. We also showcase the robustness of VOTP in the presence of visual distractors and validate its utility on real robotic tasks, where it learns meaningful rewards with minimal human input.