Scientist World: Grow a Research World for AI-Scientists
Taiming Lu, Sachin Konan, Zhuang Liu
AI 速览：作者提出将每篇论文转化为可执行环境，构建供 AI-Scientist 研究与训练的世界，并报告首批 1000 个已验证环境及初步评测与训练结果。
AI-Scientists arrived, but a world for them to grow in did not. Human scientists take theirs for granted $-$ centuries of open problems and shared ideas, waiting in the scientific literature. To AI-Scientists, that world is readable but not runnable, so we build Scientist World for them: every research paper becomes an executable environment. Our autonomous pipeline extracts each paper's open-ended research question and reproduces its proposed method as the baseline; an AI-Scientist entering the environment is evaluated against what the human authors achieved. We release the first 1,000 verified environments. With the world as an evaluation, we found that frontier AI-Scientists can already surpass the human researchers 74% of the time. As a record of how AI does science, it contains 6,400 hours of research traces, with every idea, experiment, failure, and success. As a training ground, we are able to raise (+158%) a small model to match frontier models in research ideation. At last, AI-Scientists' own world has arrived, grown from and growing with science itself.
AI 速览：提出开源多智能体系统 HEMA，用大语言模型加36个领域工具支持多轮对话式家庭能源管理，并给出基于LLM模拟用户的评测框架。
Existing home energy management systems conceptualize occupants as passive recipients of energy information and control, which limits their ability to effectively support informed decision-making and sustained engagement. This paper presents Home Energy Management Assistant (HEMA), the first open-source, multi-agent system enabling sustained human-AI collaboration - multi-turn conversational interactions with preserved context - across diverse home energy management (HEM) tasks - from energy analysis and educational support to smart device control. HEMA combines large language model (LLM) reasoning capabilities with 36 purpose-built domain-specific tools through a three-layer architecture: a web-based conversational interface, a backend API server, and a multi-agent system. The system features three specialized agents - Analysis (energy consumption patterns and cost optimization), Knowledge (educational queries and rebate information), and Control (smart device management and scheduling) - coordinated through a self-consistency classifier that routes user queries using chain-of-thought reasoning. This architecture enables various energy analyses, adaptive explanations, and streamlined device control. HEMA also includes a comprehensive evaluation framework using an LLM-as-simulated-user methodology with 23 objective metrics across task performance, factual accuracy, interaction quality, and system efficiency, allowing systematic testing across diverse scenarios and user personas without requiring extensive human subject testing. Through demonstrations using real-world household energy consumption data, we show how HEMA supports informed decision-making and active engagement in HEM, highlighting its potential as a user-friendly, adaptable tool for residential deployment and as a research platform for HEM innovation.
SAGE: A hybrid vision-language and geometry framework for work-zone pathway prediction under temporary traffic control
Jiaxi Liu, Keshu Wu, Weizhe Tang, Bofeng Cao, Sikai Chen, Xiaopeng Li, Steven T. Parker, Bin Ran
AI 速览：SAGE 将视觉语言模型（VLM）的轨迹作为对几何候选路径的偏好，在共享候选空间中分离路径构建、路径偏好与控制合规，以预测施工区临时交通管制下的通行路径。
Temporary traffic control redirects vehicles through work zones, requiring automated vehicles to connect an interpretation of the visible devices to a continuous passing path. Geometric predictors provide several plausible paths, yet selecting the one consistent with the temporary control remains difficult. We propose SAGE (Semantic-to-Action Grounded Execution), a read-decide-execute-check framework that separates path construction, path preference, and control compliance within a shared candidate space. The central idea is to use a vision-language model's (VLM's) trajectory as a preference over intact geometric candidates. A role-decoupled generator constructs the candidates, and rank fusion combines each candidate's geometric score with its agreement with the VLM trajectory. Supervision structured around the Manual on Uniform Traffic Control Devices (MUTCD) trains SAGE-VLM to state the lane status and hazard side before its passage decision and trajectory. The predicted hazard side enters image-space predicates derived from MUTCD provisions, which support a final intervention bounded in score and displacement. On 2,306 real-world ROADWork validation episodes, the candidates include a path within 10 px of the human demonstration in 87% of episodes, yet selection regret accounts for 53% of the geometric ranker's error. SAGE reduces average and final displacement errors by 8.3% and 12.0% relative to the ranker and hard-constraint violations by 37%. The trajectory opinion improves all four tested generator-selector combinations. SAGE is comparable to state-of-the-art VLM modules in accuracy, with a faster-responding VLM, and outperforms four other adapted driving VLMs in displacement error and safety measures. Project page: https://hhjj233.github.io/SAGE-page/.
He Hu, Juzheng Si, Qianning Wang, Fei Ma, Fuji Ren, Sicheng Zhao, Yucheng Zhou, Laizhong Cui
Large language models (LLMs) hold significant potential for psychological support, capable of generating empathetic responses and simulating therapeutic conversations. However, existing LLM-based approaches often lack sufficient grounding in structured diagnostic reasoning and diverse therapeutic strategies, limiting their effectiveness in complex counseling scenarios. To address these limitations, we propose PsyLLM, the first large language model designed to incorporate diagnostic- and therapeutic-oriented reasoning for mental health counseling. To develop PsyLLM, we design a novel automated data synthesis pipeline that processes real-world mental health posts collected from Reddit, where users frequently express psychological distress and seek community support. The pipeline transforms these posts into multi-turn counseling dialogues and leverages LLMs guided by international diagnostic standards (e.g., DSM/ICD) and multiple therapeutic frameworks (e.g., CBT, ACT, psychodynamic) to generate structured, clinically informed reasoning traces. A rigorous multi-dimensional filtering mechanism further improves the quality and clinical fidelity of the synthesized dialogue data. In addition, we introduce a new benchmark and evaluation protocol, assessing counseling performance across four key dimensions. Our experiments, along with evaluations by professional counseling experts, demonstrate that PsyLLM consistently outperforms various baseline models on this benchmark. Code, data, and models are available at https://github.com/Emo-gml/PsyLLM.
Enabling vision-language models (VLMs) to control humanoid robots requires both reliable whole-body interfaces and knowledge of how to use the body that carries over across tasks. We present HITCH (Harness Interfaces To Command Humanoids), a harness in which perception returns the 3D position of any point the VLM selects in its image, and execution turns commands for the pelvis, torso, hands, walking, and gaze into whole-body motion with a learned tracker. We find that both measuring the scene and executing commands accurately raise task success. We then fine-tune a VLM on trajectories in which teacher policies solve five base tasks through HITCH, so that it learns how to operate the humanoid. Without task-specific trajectories, the fine-tuned model completes 51% of SIMPLE episodes and 42.5% of HumanoidArena episodes, and adding five trajectories per SIMPLE task raises its SIMPLE success to 95%. On a real G1, it carries a box and walks to a goal without real training data.
Conceptual abstraction and gender are two equally central aspects of the human mind, yet they have traditionally been studied in isolation from one another, as though they were fundamentally separate. Here, we find gender and abstraction are systematically linked throughout the space of concepts spanning single (N = 40,000) and multi-word expressions (N = 67,000): across billions of words and millions of images on the internet --- and across numerous measurement techniques, including human and AI classifications --- more abstract concepts are more associated with men, a pattern we show replicates in languages around the world. This male abstraction bias not only holds across semantic domains, but it also maps onto grammatical structure: verbs and adverbs are more abstract and more male-coded, revealing a system-wide trace of stereotypes accentuating agentic qualities in men. Furthermore, we show this bias readily emerges in the free associations humans and AIs produce in controlled online tasks, indicating its psychological internalization and its algorithmic reproduction. These findings identify a latent structure in conceptual space that unifies a range of stereotypes: for example, framings of women as more embodied, emotional, and warm, versus men as more cognitive, rational, and competent; as well as erroneous views of men as better at mathematics, science, and engineering. More broadly, by analyzing an unprecedentedly large, multimodal sample of online data, we provide culture-wide evidence of gender as a core organizing principle shaping conceptual space far beyond the characterization of people and social roles.
Hong-Su Nam, Seokhwan Lee, Jungwoo Cho, Hak-Tae Lee
A sequential engineering framework is presented for urban drone delivery, linking hub location, route planning, mission-specific vehicle sizing, and energy and operating cost assessment. Aggregate origin-destination demand is assumed known a priori. For each hub configuration, the maximum required link distance is minimized, and feasible routes that return to their departure hubs are generated. Route frequencies are determined using mixed-integer linear programming to satisfy delivery demand. The resulting range requirement guides motor-propeller selection and battery sizing to minimize vehicle takeoff mass. Flight distances and frequencies, cruise performance, and vehicle dimensions are combined to estimate monthly energy consumption, fleet size, hub area, staffing, and operating cost. The framework is applied to district-level Seoul delivery data. Increasing the number of hubs from one to four reduces the required range from 30.6 to 21.9 km and vehicle takeoff mass from 8.12 to 6.12 kg. Despite a 21.5% increase in total flight distance, monthly flight energy consumption decreases by 13.1%, from 15.2 to 13.2 MWh. Battery replacement is the largest vehicle-related operating expense, while operating cost per delivery increases from USD 1.18 to USD 2.20 because of higher rent and labor costs at the hubs. Among the designs evaluated, range-matched vehicles achieve the lowest operating cost in the two- and four-hub networks; a longer-range vehicle is least costly in the three-hub network. Sensitivity analyses further examine the effects of delivery demand and operating assumptions. The results demonstrate that reductions in vehicle mass and energy consumption do not necessarily lead to lower system operating cost.
Seungyong Lee, Suhyeok Jang, Minyoung Kim, Jinwoo Shin
Robot foundation models have shown remarkable performance in robot control, yet adapting them to a new task still relies heavily on teleoperated demonstrations. Egocentric human videos offer a scalable substitute, but overlaying rendered robots on human videos or retargeting human poses into robot actions leaves significant visual or action gaps for high-DoF robot embodiments. To address this limitation, we propose RoboTryOn, which treats human-to-robot translation as virtual try-on: try-on keeps the person and swaps the garment, RoboTryOn keeps the demonstrated motion and swaps the human for the robot. Specifically, RoboTryOn converts a human video into a video of the target robot specified by a reference image, while preserving the original task, objects, and background. To maintain visual consistency, we translate selected keyframes autoregressively and use them to guide video synthesis with motion conditioned on the human hand skeleton. Additionally, we use an inverse dynamics model (IDM) that incorporates depth features to infer action labels from synthesized videos for policy learning. Our experiments across multiple humanoid embodiments show that policies trained with our synthetic data improve the average success rate on novel tasks, including challenging bimanual manipulation, from 2% to 60%, without any real robot demonstration of those tasks. Moreover, training with our synthetic data matches the performance of policies trained solely on teleoperated demonstrations at approximately a third of the collection cost, demonstrating the cost efficiency of our approach.
Dense labels of the aorta in CT are slow to draw, and the public aortic CT datasets that could stand in for them each label different structures under their own conventions. Large language model (LLM) agents can now write code and read images. We asked whether such a model labels these scans better as a free-form agent that writes its own segmentation code, or inside a harness, where an automated pass labels the scan and the model reviews the result. We built a harness, Taurus, and compared it with free-form agents by blind label recovery: we withheld the human labels of six public datasets (939 volumes) and scored every approach against each dataset's own classes. In the first version of Taurus, the model's review raised the primary metric on only two of six datasets, by less than the noise between two reviews of the same case. A free-form agent running Claude Opus 5.5 scored higher on most datasets. Moving what the free-form agents did well into the automated pass then carried most of the gain on 38 held-out cases. The final version smooths the label surfaces with a fixed rule and checks for the large errors the automated pass tends to make, such as an aneurysm drawn to half its size. Each check that fires puts a multiple-choice question to the model, with candidate corrections drawn on the scan. It matched the surface smoothness of the human references, and in the primary metric we found no difference from the free-form Opus agent, at under half its cost, $1.22 against $2.71 per case. Most of what made Taurus competitive came from the harness and its automation rather than from the language model itself. Taurus shows that this kind of annotation can be automated, so that expert time can move from drawing labels to checking them.
This paper presents a layered control framework that recasts high-dimensional system dynamics as a composition of dynamically interconnected subsystems, with dynamic dual-arm manipulation on the humanoid serving as a representative application. The humanoid is decomposed into a single rigid body (SRB) representing the torso and floating-base dynamic models representing the articulated arms and lower-body. The rigid connections between arm and torso and the closed kinematic chain formed between the end-effectors during cooperative manipulation are represented as holonomic constraints. The corresponding Lagrange multipliers parametrize the interaction wrenches transmitted at the shoulders and between the end-effectors. The high-level nonlinear model predictive control (NMPC) jointly optimizes the subsystem state trajectories, arm torques, ground reaction forces (GRFs), and interaction wrenches, enabling online wrench design and optimization while preserving the dynamic coupling among the independently modeled subsystems. The optimized arm torques are applied directly for cooperative dual-arm manipulation, whereas the prescribed GRFs are mapped to lower-body joint torque for whole-body coordination. We validate the proposed framework through numerical simulations and hardware experiments on a G1 humanoid robot. The results demonstrate optimal interaction wrenches consistent with simulated contact measurement, cooperative dual-arm trajectory tracking, and stable whole-body coordination while executing the NMPC in online manner.
Power matters for AI safety. Optimal agents tend to acquire power across a wide range of objectives (instrumental convergence). As we embed AI agents into socio-technical systems, humans risk gradual disempowerment. Both concerns call for a formal, quantitative measure of power. Two such measures in the context of Markov games are information-theoretic empowerment, the capacity of the channel from an agent's actions to future states, and capability-based individual power, a measure of goal attainment built from the probabilities that a human would attain each goal in a family of candidates were they pursuing it. We show where the two meet. From individual power, Fano's inequality yields a floor $B(s)$. This floor is a lower bound on what we term goal revelation, the mutual information $\mathbb{I}(G;ξ\mid s)$ between the human's latent goal and the realised trajectory. Empowerment caps revelation from above, so that $B(s) \le \mathbb{I}(G;ξ\mid s) \le \mathbb{I}(\mathbf{A}_h \to ξ\mid s) \le \mathfrak{E}_h(s)$. Each quantity can be large while the one below it stays at zero, highlighting the following failure modes. An assistant maximising empowerment can prefer an interface offering broad influence over goal-irrelevant degrees of freedom to one that lets the human attain every goal they might hold. An assistant maximising revelation cannot tell a working interface from one rewired so that every goal yields the wrong outcome, since relabelling goals leaves mutual information unchanged while zeroing every attainment probability. Positive guarantees therefore come only from the attainment end of the chain. A satisfaction floor, computable from two statistics of interaction logs, offers a way to monitor the power users retain in a deployed system, and simulations in random environments show the three quantities are correlated but far from interchangeable.
Jeffrey Lee, Geoffrey Hanks, Evangeline Bouganim, Yue Hu
Robotic automation of laboratory tasks has become a growing interest to remove researchers from hazardous environments and repetitive procedures. However, even equipment as simple as squeeze bottles poses a challenging manipulation problem due to its deformability. This paper presents an adaptive hybrid vision-force controller for robotic grippers to effectively manipulate laboratory squeeze bottles to dispense liquid. This approach integrates force and vision feedback according to the interaction state, enabling effective squeezing while accounting for bottle deformation and recovery. The proposed controller is validated by comparing its performance with a pure force controller on 12 combinations of conditions consisting of 2 target dispense volumes, 2 plastic squeeze bottles, and 1 to 3 squeezes used to dispense. Human demonstrations of these operating conditions were collected using electromyography (EMG) sensing mapped to a desired grip force, which was then used as the force reference for the controllers to squeeze the bottles. Results demonstrated that the proposed adaptive hybrid vision-force controller achieved more accurate liquid dispensing with a mean absolute error (MAE) of 6.54 mL compared to a MAE of 15.92 mL from the pure force controller, while also replicating the human action recorded from the EMG signal more closely.
Backward search over the run-length compressed Burrows-Wheeler transform (RLBWT) of a pangenome takes one step per character of the pattern, even when long stretches of the pattern match material that occurs many times in the collection. Two-level indexing also indexes a parse of the collection, so that a search can take one step per phrase. We parse at closed syncmers, whose phrases are prefix-free, and add blocks of continuations: when each phrase of a stretch of the pattern is the smallest phrase that follows the previous one anywhere in the collection, the search crosses the whole stretch in one step, on the BWT of a sequence of block symbols. We prove that the three BWTs stay in the same order, so the search still returns exactly the BWT interval of every match, and we bound the extra space by terms that depend on the length of the collection only through the number of distinct closed syncmers. On chromosome 19 haplotypes from the Human Pangenome Reference Consortium, finding the maximal exact matches of length at least 1000 of two haplotypes, on both strands, takes 5.5 to 5.7 seconds with 16, 32 or 64 other individuals, while backward search on a run-length compressed suffix array of the same collection takes 40 to 60 seconds and grows with the collection. We compare this with suffix-tree path decompositions (STPDs): over the parse, an STPD is no faster and needs about twice the memory; over the text it is about 1.7 times faster but needs the whole text in memory, about five times the space; and the published STPD-index, which compresses its text with relative Lempel-Ziv, is 8 to 10 times slower. Keeping a toehold through all three kinds of steps lets the search also list every occurrence of every MEM, for about 2 more seconds.
Backward search over a run-length compressed Burrows-Wheeler transform (RLBWT) takes one step per pattern character, even when long stretches of the pattern match material that recurs throughout a highly repetitive collection. We describe shortcuts: for a precomputed set of pairs $(α,β)$ of strings, a backward search that has just matched $β$ and finds $α$ immediately to its left in the pattern can cross $α$ in one step. Since a shortcut is taken only after its pair has been recognized in the pattern, the search returns a report of the shortcuts it took along with the usual BWT interval. Our motivation is pangenome graphs, where such pairs mark the places where a pattern passes through a vertex, so the report gives the path a match takes through the graph. We work out the case of a single linear reference $R$, with $β$ ranging over the length-$\ell$ windows of $R$ starting at multiples of $\ell$, and $α$ over the $2^t \ell$ characters preceding such a window. We show that the search counts correctly, that under stated conditions it crosses a stretch of the pattern matching $m$ consecutive windows of $R$ with $O(\log m + m / 2^τ)$ shortcuts, where $2^τ\ell$ is the longest shortcut, and that the shortcuts can be built efficiently from the RLBWT. On the Pizza&Chili cere collection of 37 genomes, shortcuts make backward search 2.3 to 4.2 times faster on exact patterns of length 200 to 1000, and 1.7 to 2.0 times faster with 1% of the characters substituted, for 86% more space. On 11 gigabases of genomes of seven bacterial species, they make it 1.6 to 1.9 times faster on exact patterns for 32% more space. On whole chromosome 19 haplotypes against 16 to 64 others from the Human Pangenome Reference Consortium, shortcuts in both directions make Boyer-Moore-Li search for maximal exact matches of length at least 1000 about 8 times faster.
Data scarcity limits inference in many scientific and policy domains. Survey data are essential for decision-making, but sparse samples often fail to capture fine spatial granularities. We evaluate normalizing flows, a generative model that learns complex data distributions and can be conditioned on exogenous contextual features, in controlled data scarcity scenarios. Across eight household survey datasets spanning six low-income or middle-income countries in the humanitarian domain, we show that context-conditioned generative models can refine sub-national survey distributions under severe data scarcity, and that performance increases systematically with the richness of the conditioning information. These findings support a general principle for survey data augmentation: generative models can improve sub-national estimates when the sparse sample retains sufficient support and contextual covariates encode relevant local heterogeneity. By learning full conditional distributions rather than point estimates, the approach provides fine-grained evidence for humanitarian decision-making and resource allocation.
Marcin Moskalewicz, Anna Sterna, Karolina Drożdż, Kacper Dudzic, Marek Pokropski, Paula Flores
Building on a human-led thematic analysis of clinical life-story interviews (> 150,000 words) with inpatients with Borderline Personality Disorder, this study examines the capacity of large language models (OpenAI's GPT, Google's Gemini, and Anthropic's Claude) to support qualitative clinical analysis. The models' interpretative potential was evaluated using a mixed-methods approach. Study A involved blinded and non-blinded judges in phenomenology and clinical psychology. The experts assessed the validity of AI-generated content using semantic congruence, Jaccard coefficients, and multidimensional validity ratings, including credibility, coherence, the substantiveness of results, and grounding in qualitative data. In Study B, neural methods were used to embed human- and model-generated theme descriptions in a multidimensional vector space. This approach provided an objectified computational measure of the difference between human and model semantics and linguistic style. In Study C, complementary non-expert evaluations were conducted (N=115) to examine the influence of thematic verbosity on the perception of human authorship and content validity. Overall, the results of AI analysis showed a highly variable overlap (0-58%) with the human interpretation, while all models identified themes originally omitted by human researchers, proving their capacity to mitigate human bias. At the thematic level, external evaluators were unable to reliably distinguish human-authored themes from those generated by AI. In terms of content validity assessed against raw data by high-level experts in the blinded mode, the performance of Gemini 2.5 Pro was indistinguishable from that of humans. The comparison of semantic vector embeddings showed that its style was also the closest to humans.
Evanfiya Logacheva, Arto Hellas, Tsvetomila Mihaylova, Juha Sorva, Ava Heinonen, Juho Leinonen
We investigate how provision of in-context examples based on discourse similarity affects quality and hallucination in diagram code generation from educational materials. Using a sample of texts as sources for diagram generation, we test two methods using a conventional text-to-diagram task format, instruction-based zero-shot generation and ICL with text-to-diagram examples, and an unconventional ICL method with a discourse analysis-to-diagram example, which requires a model to generate a diagram from a discourse analysis. We evaluate AI-generated diagrams for faithfulness to provided context and their quality outcomes using both human and automated evaluation. Bayesian data analysis of the human diagram evaluation shows that there is significant evidence that the unconventional task format improves context faithfulness in diagram generation compared to the zero-shot instruction with reducing the odds of context faithfulness hallucination by nearly $70\%$; however, we do not find the ICL having a significant effect on the diagram aesthetic presentation and superficial coherence. Evaluation with the LLM-as-a-judge G-Eval metric shows nearly 4.5-fold increased odds of the perfect faithfulness outcome for the novel ICL task. Our study provides a novel insight into ICL format effect on superficial fluency and faithfulness in context-grounded structured diagram code generation.
Benchmarks for visualization agents score the agent alone; the human only supplies reference answers. For high-dimensional data this misses what matters: the knowledge of which variables count is held by on-site experts, not by the model, so the useful question is how well an analyst and an agent do together. We study one specific form of collaboration, the agent proposes a view, the analyst accepts, rejects, or modifies it, and those judgments are kept for later sessions, and propose five measures for it: complementarity, intervention efficiency, appropriate reliance, knowledge retention, and agency. The first four are computed from interaction logs alone. We test the measures, not people: in a testbed for parallel-coordinates design on public manufacturing data (UCI SECOM), with a simulated analyst whose behavior we control, all four are computable and respond to the controlled differences. Whether they reveal anything about real analysts is the next step; we release the testbed for it.
Can independent components be recovered when all of them are Gaussian? Yes, if their variances change across environments; but how many environments, and how many samples as the number of components $d$ grows? We answer both for the joint diagonalization of $K$ covariances. Each component has a signature, its variances across the environments. Separating component $j$ costs the Riesz 2-energy $E_j$ of this signature, a sum, over the other components, of inverse pairwise Fisher informations. For Gaussian sources no estimator in $O(d)$ beats $cE_j/(n+E_j)$ at any $n$ and $d$, and above an explicit sample size a spectral estimator attains $E_j/n$ up to a constant, for every mixing and without the signatures (given a lower bound on the pairwise information). Since signatures live in $K-1$ dimensions, the rate obeys a dimension law: $d^2/n$, $d\log d/n$ and $d/n$ for $K=2$, $3$ and $K\ge4$ well-spread environments. What counts is the dimension the signatures span, not their number, and $\log d$ well-designed environments make $d\log d$ samples suffice. Projecting the covariances onto one linear combination pays $d^2/n$ at first order instead. In simulations up to $d=1000$ the predictions hold and joint diagonalization with $K\ge4$ attains the rate once $n/d$ is in the hundreds; on task fMRI the energy says which components are hard to separate.
Michikuni Eguchi, Shigeo Yoshida, Mai Nishimura, Takefumi Hiraki
Swarm robots have attracted growing interest as a physical interface for human--computer interaction, since a swarm can reshape its formation to fit the objects and surroundings it works with. For a human operator, however, adapting the swarm's behavior remains difficult. Most existing methods require the user to specify where the swarm should go, and they offer no intuitive way to change the collective's behavior as each task demands. We present Fluidic Body, a system that lets a person operate a swarm of fluid-imitating robots as an extension of the hand. The system places a virtual fluid between the hand and the swarm, and the user adapts the swarm's behavior to each task by changing the fluid's properties through simple hand movements. When the user opens the hand, the swarm becomes soft enough to deform and flow around obstacles. Closing the hand turns it into a hard, cohesive body that can push objects. We evaluated the controller's performance and its operability in two simulation experiments and one demonstration with physical robots. First, our controller handled both obstacle avoidance and object transportation better than an existing method that behaves the same way in every task. Second, in a VR user study, 17 participants completed the combined avoidance-and-transportation task faster and more reliably than with that method. They also reported a lower workload and a stronger sense of agency. Finally, this experience carried over to physical robots in a real-world demonstration (88 participants). These results suggest that treating a swarm as a fluid, a material people already understand, allows a robot swarm to work as a deformable extension of the human body.
Yusheng Dai, Vladimir Iashin, Kangdi Wang, Yuxuan Jiang, Ilpo Viertola, Weiqiang Wang, Qiuhong Ke, Min Zhao 等 9 位作者
Streaming video-to-audio (V2A) generation is essential for immersive interactive applications. However, real-time constraints leave existing streaming V2A methods far behind offline models: (1) visual frontends shrink their receptive fields to cut look-ahead latency, severely degrading audio-visual alignment; (2) causal autoregressive diffusion backends denoise each audio patch in isolation, impairing audio quality. To this end, we present LiveFoley, a streaming causal autoregressive diffusion architecture that addresses both bottlenecks with imperceptible latency. For the visual frontend, we introduce a Trailing Sliding Window that converts non-causal segment-level encoders into streaming modules without retraining. By simply realigning the window, it preserves precise audio-visual alignment, while a configurable hop size reduces the frontend look-ahead well below the human perceptual threshold. Building on this tunable hop size, we propose the Unified Varhop Framework to accommodate scenarios with diverse latency and computation budgets. For the audio backend, we explore widening the denoising window of the local diffusion transformer (LocDiT) head in two temporal directions. History LocDiT adds past context to improve quality at no latency cost and serves as our default. Alternatively, Future-Warmup LocDiT breaks the autoregressive deadlock with an approximate future condition, further improving generation quality at the cost of an acceptable backend look-ahead. Extensive experiments demonstrate that LiveFoley matches offline models in temporal alignment and leads streaming systems in generation quality. Project page: https://dalision.github.io/livefoley-page/
Manipulating diverse objects with multi-fingered dexterous hands is challenging due to the high dimensionality and complex dynamics. Human-Object Interaction (HOI) datasets provide rich knowledge about task information and embodied interactions. Instead of solely imitating the human demonstrations, our method learns to holistically predict future hand-object states by leveraging these datasets. The predicted future states of the object can serve as a general-purpose reward term for reinforcement learning, reducing reliance on task-specific reward engineering and enhancing generalization across tasks. We conduct extensive experiments on three manipulation tasks in simulation and the real world. Our approach outperforms existing SOTA methods in both success rate and generalizability on novel objects. Furthermore, we validate the cross-embodiment compatibility of our methods by successfully deploying the skills on different robot hands.
Keonhee Jang, Kanghoon Lee, Seokwoo Yun, Namwoo Kim, Yoonjin Yoon
Pedestrians increasingly encounter robots in urban public spaces without interacting with them, but how robots change the perceived safety of a street remains unclear. We examine how robot presence and type affect perceived safety across urban contexts. We generate 18 images of Seoul streets that vary in time of day, road class, and robot condition (no robot, a delivery robot, or a humanoid). Two multimodal large language models (MLLMs), Qwen3.8-27B and Gemma4-26B, each conditioned on 500 synthetic personas, choose the safer image in each pair, and 25 human respondents complete the same task. Both models prefer streets without robots and favor the delivery robot over the humanoid. Robots are chosen more often at night and less often by older personas, whereas road class has smaller effects. In the model rationales, the delivery robot appears as a physical obstacle, whereas the humanoid appears as either a threat or a source of protection. Like humans, both models chose robots more often at night, but chose the humanoid over the delivery robot far less often, and Gemma showed a night effect for the humanoid about three times that of humans. These results show that the effect of a robot on perceived safety depends on when it appears, and that persona-conditioned MLLMs capture only part of human judgments.
Nicole Vasilevsky, Aleix Puig-Barbe, Andreas Bueckle, Ellen M. Quardokus, Yongxin Kong, Divya Prasanth Paraman, Jie Zheng, Yongqun He 等 11 位作者
The Human Reference Atlas (HRA) v2.5 includes 4,807 anatomical structures, 1,347 cell types, and 2,215 biomarkers (genes, proteins, lipids); 81 organs with 1,392 anatomical structures modeled in three dimensions (3D). Mapping new data into the HRA requires 3D spatial registration and/or cell type annotation (CTann). Many CTann tools exist but few use cell type names linked to Cell Ontology (CL), making comparisons across tools, tissues, and organs difficult. The HRA supports five transcriptomics CTann tools (Azimuth, CellTypist, popV, FR-Match, Pan-Human Azimuth), four sc-proteomics CTann tools (DeepCell Types, DeepCell Types-HuBMAP, Robust Image-Based Cell Annotator [RIBCA], and SpaTial cELl LeARning [STELLAR]), plus a Cell Distance Explorer (CDE) Spatial Omics collection covering previously published annotations across 12 tissue types with 47,349,496 cells. Assigned cell types differ in number, resolution, and coverage. This paper details how 10 crosswalks were constructed, validated, and used to compile a "Supertree" that can be used to compare CTann across organs and technologies in support of reference atlas constitution and usage. All data and code can be accessed at https://cns-iu.github.io/hra-ctann-supporting-information.
When does an AI assistant improve human decision-making, and when does it hurt? We argue that the answer depends not just on AI accuracy but also on how much its evidence overlaps with the human's and how the human combines the two. A worker makes a binary decision by observing her own signal and receiving a recommendation from an AI. We develop a human behavioral model with two primitives: her awareness of the overlap and her trust in the AI. Our analysis gives a closed-form phase diagram over AI capability and overlap, with regions of complementarity, impairment, and automation. A worker who neglects the overlap (correlation neglect) can be made less accurate by a more capable AI, double-counting the shared evidence. Trust reshapes the regions - undertrust may compensate for correlation neglect, so that the trust level maximizing accuracy depends on the deployment parameters. Empirically, we identify the two model parameters by pairing the same workers with different AIs, and break the accuracy loss into a correlation-neglect cost and an undertrust cost, one tied to reducing overlap on the AI side, the other to calibrating trust on the human side. On ChatBench, AI assistance raises accuracy by 21 points, yet workers apply only about a quarter of the Bayes-rational weight to AI information; GPT-4o falls in the automation region and Llama-3.1-8b stays at the complementarity boundary; and calibrating trust recovers about 2.6 points for GPT-4o but almost nothing for Llama-3.1-8b.
Temporal-difference (TD) learning underlies many prediction and control algorithms, but can diverge when combined with off-policy learning, bootstrapping, and function approximation. TD with Regularized Corrections (TDRC) is theoretically sound and practically simple, but existing convergence theory covers only the one-step case under i.i.d. sampling, while the Huber Bellman-error formulation has not been extended to TDRC or eligibility traces. In this work, we establish convergence of constrained TDRC($λ$) with diminishing stepsizes and non-i.i.d. off-policy data, for state-dependent and history-dependent $λ$-parameters. We also propose TDRC($λ$)-Huber, which preserves the TDRC($λ$) fixed point while reducing sensitivity to large Bellman errors. For both algorithms, we apply an ODE-based weak-convergence method from stochastic approximation theory. For the nonlinear mean dynamics of TDRC($λ$)-Huber, we give an integral small-gain condition that guarantees uniqueness and global asymptotic stability of the target equilibrium. We then establish convergence of the constrained algorithm, derive computable sufficient conditions for convergence, and identify a regularization-based convergence mechanism shared by linear TDRC($λ$) and TDRC($λ$)-Huber. When the mean TD matrix is positive stable, both constrained algorithms converge under sufficiently strong regularization without requiring positive definiteness.
Diffusion and flow-matching based text-to-speech (TTS) models excel in naturalness but often lack explicit emotion control, as emotional signals remain entangled with speaker identity. We discover that emotion embedding emerges as a linearly decodable direction of frozen hidden states, nearly orthogonal to the direction embedding speaker identity. This inspires a plug-and-play framework DUET for emotion control over pretrained diffusion and flow-matching based TTS models. During generation, DUET unifies dual-space control to achieve fine-grained emotion intervention in a single per-step update: hidden space steering shifts generation along the target emotion direction, while mel-space guidance refines spectral details through gradients backpropagated from a differentiable vocoder. We validate DUET on five architecturally diverse pretrained TTS backbones across three datasets, where it outperforms 10 supervised state-of-the-art emotional TTS baselines across paradigms and achieves the highest human-rated emotion appropriateness. To further showcase its qualitative behavior, we deploy DUET on an Ameca humanoid robot, where it produces richly expressive emotional speech on the humanoid, demonstrating the strong potential for plug-and-play affective interaction for embodied agents.
Office documents are now routinely uploaded to large language models (LLMs) for processing. However, an LLM does not always see what a human sees, because it typically reads the text that an extractor produces from the document's underlying Office Open XML (OOXML) files. Depending on the implementation, extractors differ in which OOXML parts they read and how they interpret them, so their output can differ from what Office applications display. For example, by slightly shifting the positions of text boxes on a slide, an adversary can swap the order in which the model reads them while the slide looks unchanged. We show that this extraction divergence is an attack surface for LLM applications. In this work, we first measure what an adversary can exploit by analyzing the widely used extractors that LLM applications rely on. By combining static source code analysis with dynamic analysis, we identify 19 OOXML channels where extractors read content a human cannot see and 12 where they drop content a human can. We also find that widely used extractors ignore display formatting and reading order, so even visible content reaches the model in a different form. Building on these findings, we construct five attacks that lead the model to interpret a document differently from a human reviewer, without injecting any instruction. We test the attacks on four major chat assistants (ChatGPT, Claude, Gemini, and Microsoft 365 Copilot) and four coding agents (Codex, Claude Code, Antigravity, and GitHub Copilot CLI), and every one of them is misled by at least one attack. Finally, we propose Prism, a lightweight defense that reconstructs the human view from the OOXML specification before passing the text to the model. Prism also generalizes to unseen attack payloads.
LLM-as-a-judge is a common substitute for costly human evaluation, and aggregating several judges is the standard remedy for any single judge's bias. Judge pools mix judges of very different reliability, while the human-labeled sets that validate them hold a few hundred items or fewer. In this regime the per-item mean weights weak judges as heavily as strong ones, and confusion-table models such as Dawid-Skene must estimate a per-judge error table from too few items. We propose AGORA (Aggregation Of Reliability-weighted Assessments), an item-response-theory aggregator that places every item on one latent quality scale and describes each judge by a sensitivity parameter and ordered rating thresholds, so judge reliability is learned from all items jointly and without human labels. Across three benchmarks, AGORA matches or exceeds Dawid-Skene on large evaluation sets, and its advantage widens as the set shrinks toward a few hundred items. It remains accurate as pools grow more uneven, a label-free dispersion statistic indicates in advance when trust weighting helps, and its learned trust weights track per-judge human validity.
Multimodal large language models (MLLMs) allow social scientists to analyze video at scale, but researchers must decide which parts of a video to provide as input. We use the term representation validity to examine how text, audio, video, and combined inputs affect measurement and subsequent analysis. We compare input representations across six open-weight MLLMs and three human-annotated video datasets. No representation consistently performs best. On seven news-related constructs, native-video performance varies little across models (macro-F1 = .70-.77), while extracted-text performance ranges from .52 to .80. Three models perform better with text and three with video. Combining inputs rarely improves performance. Higher classification accuracy also does not reliably produce estimates closer to those based on human codes. The highest-F1 representation best reproduces human-coded regression coefficients in 17 of 42 comparisons. Even when each model uses its highest-F1 deployable representation, 45% of sample-prevalence estimates differ from the human-coded benchmark by more than ten percentage points. Input representation therefore requires validation for both the construct being measured and the quantity estimated from the resulting labels.