Person identification from millimeter-wave (mmWave) radar point clouds has primarily focused on gait. However, since indoor walking is typically brief and fragmented, gait-based identification is largely constrained to designated passages that support continuous walking. Meanwhile, non-gait activities of daily living (ADLs) are pervasive and also encode discriminative personal signatures. In this study, we evaluate the feasibility of ADL-based identification from sparse mmWave point clouds using mm-ADL, a dataset comprising seven core ADLs collected from 11 subjects under controlled protocols. Moving beyond uniform gait introduces significant activity-dependent variations that confound identity representations. To address this challenge, we systematically benchmark three activity-identity learning paradigms: monolithic learning, joint multi-task learning, and two-stage activity-conditioned routing, under both closed-set identification (ID) and subject-disjoint re-identification (ReID) biometric settings. In parallel, we propose Dual-Stream Static-Dynamic PointNet (DS-SDPNet), an architecture tailored to radar point clouds that captures complementary spatial morphology and motion dynamics. Across seven ADLs, DS-SDPNet achieves 70.2% mean closed-set identification accuracy and outperforms existing baselines on both mm-ADL and mmGait. Moreover, benchmarking the three paradigms without ground-truth activity labels during inference yields up to 68.0% closed-set accuracy and 75.4% ReID mAP, providing a systematic baseline and confirming the feasibility of point-cloud-based biometric identification across diverse daily activities.
Robots increasingly run large vision-language-action policies on remote servers, making the camera uplink the bottleneck. Existing video codecs spend bits on content the receiver could predict: the robot's own motion. We introduce the kinematic hypothesis context, a zero-rate conditioning signal that both ends build from joint states, the robot model and the camera calibration. Ego-motion homographies, per-link robot reprojections, an end-effector plane and a background memory warp the previous reconstruction into hypotheses that condition a learned video codec. We prove that history-only and global-pose conditioning with a limited receptive field cannot access this information under large motion, and that coding gains become task gains at the policy's success cliff. On LIBERO, the context saves 23.9% bit rate over a capacity-matched history-conditioned codec and 40-56% over x264, x265 and SVT-AV1, and needs 23.2% fewer bits than DCVC-RT at equal policy action deviation; trained to lower rates, it matches DCVC-RT in PSNR. In closed loop at equal rate, OpenVLA-OFT success nearly doubles from 22.8% to 43.1% at 6 kbps (72.6% with lower-rate training, 23.6 points above DCVC-RT), pi_0.5 success rises from 55.5% to 77.4%, and the gain grows under packet loss. The context transfers across three codec hosts and to real robot data.
Scientific experiments are increasingly selected, executed and interpreted by learned systems that act through physical interfaces. Robot learning builds physical competence, whereas AI for science chooses which experiment to run, each reporting different endpoints. This review treats their interface as one problem, embodied AI for science, where the experiment itself is partially observed: specimen state, apparatus state and measurement processes are latent, actions are costly and irreversible, and missingness depends on execution and outcome. We organize the literature by three closures operating at distinct levels: operational closure (the experiment realized is the one intended), epistemic closure (the result updates beliefs or selects subsequent experiments) and evidential closure (the record supports a claim). We analyze workflows through a record-linked evidence cycle - Record, Diagnose, Calibrate, Confirm - as analytic functions with feedback. A taxonomy crosses these closures with six delegated decisions and five embodiment classes. Our formulation separates commanded, realized and observed conditions and models diagnosis with explicit resource, risk and sample-change costs. We compare experiment-selection methods by terminal decision, review perception, protocol compilation, visuomotor policies, world models, recovery and transfer by evidence emission requirements, and compare chemistry, materials, biology, microscopy and field workflows by what their records establish. An appendix provides an evidence map of 28 workflows (26 physical, two software-boundary), characterizing validation pathways and claim confirmation requirements. We propose evaluation that separates actions, valid experiments, campaigns and confirmed claims, framing open challenges as controlled comparisons with measurable failure conditions.
Siyang Wu, Weizhi Zhang, Chen Yang, Esther Xu Fei, Jorin Overwiening, Xu Pan, Jing Tang, Yixuan He 等 37 位作者
Scientific agents collaborate with researchers whose expertise and preferences differ, but benchmarks usually hold the collaborator fixed. We introduce PersonaScienceBench: 200 tasks from five benchmarks across four scientific domains. Simulated researchers provide context and feedback according to persona profiles, while task objectives and grading remain fixed. Agents encounter researcher preferences only through conversation. Across eighteen modelconfigurations, task performance, fact recovery, and self-repetition reveal different strengths. Detailed GPT 6 Astra analyses show distinct agent responses across researcher profiles; interaction traces illustrate targeted clar ification and workflow revision. These behaviors can differ even when final scores match. PersonaScienceBench therefore evaluates how agents collaborate with diverse researchers alongside whether they complete the task.
Embodied systems can update policies, world models, skills, and executable programs from deployment experience, allowing one interaction to benefit later tasks. When a system also revises and reuses its data-acquisition, feedback, or learning procedures, experience changes the process that produces capability, introducing recursive self-improvement. This paper surveys research in physical AI through the connection between persistent capability and its improvement process. We first explain the functional distinction between learner updates and improvers, then compare policy learning, imagined experience, memory and skill accumulation, and executable harnesses. These approaches connect to revisions of curricula, rewards, and research programs in different ways. Contact validity, recovery, retention, and resource costs determine whether the resulting gains accumulate during continued deployment. An L0-L3 rubric distinguishes task-local adaptation, persistent capability gain, inherited process revision, and verified improver gain. Representative benchmarks and within-study contrasts relate these distinctions to existing results. We further propose common-data and interactive comparisons that evaluate reference and revised improvers from shared learner starts, separating the use of experience from its acquisition. This perspective connects the diverse methods, physical operating conditions, and evaluation evidence, underpinning embodied RSI. A comprehensive list of studies in this survey is available at https://github.com/nkd-lkz/physical-ai-rsi-survey.
Freja Thoresen, Aidan Cowley, Romeo Haak, Jonas Lewe, Clara Moriceau, Piotr Knapczyk, Victoria S. Engelschiøn
Human exploration of the moon is expected to resume in the next decade, following the last such activities in the Apollo programme time. One of the major objectives of returning to the Moon is to continue retrieving geological samples, with a focus on collecting high-quality specimens to maximize scientific return. Tools that assist astronauts in making informed decisions about sample collection activities can maximize the scientific value of future lunar missions. A lunar rock classifier is a tool that can potentially provide the necessary information for astronauts to analyze lunar rock samples, allowing them to augment in-situ value identification of samples. Towards demonstrating the value of such a tool, in this paper, we introduce a framework for classifying rock types in thin sections of lunar rocks. We leverage the vast collection of petrographic thin-section images from the Apollo missions, captured under plane-polarized light (PPL), cross-polarised light (XPL), and reflected light at varying magnifications. Advanced machine learning methods, including contrastive learning, are applied to analyze these images and extract meaningful features. The contrastive learning approach fine-tunes a pre-trained Inception-Resnet-v2 network with the SimCLR loss function. The fine-tuned Inception-Resnet-v2 network can then extract essential features effectively from the thin-section images of Apollo rocks. A simple binary classifier is trained using transfer learning from the fine-tuned Inception-ResNet-v2 to 98.44\% ($\pm$1.47) accuracy in separating breccias from basalts.
Effective collaboration requires language models to understand what others know and believe, and to act accordingly. Models that assume others share their information may omit necessary communication and undermine coordination. Many theory-of-mind (ToM) benchmarks assess mental-state inference through story questions, but correct answers do not establish that models can use these judgments in subsequent tasks. We introduce a multi-level evaluation built on existing, verifiable tasks. We extend household collaboration and calendar scheduling with asymmetric information while retaining their original constraints and validators. Independent interfaces assess world understanding, information access, beliefs, decisions, and execution. Controls supply personal evidence or correct beliefs to distinguish state recovery from state use. Across six model configurations, these abilities diverge. On matched scheduling cases, Claude recovers all agent beliefs correctly but produces subjectively optimal plans in only 67.9% of cases. In household collaboration, supplying beliefs raises Qwen-27B success from 55.0% to 75.0%, yet Qwen-4B answers every fixed supplied-belief source question correctly while completing only 20.0% of the corresponding continuation condition. These results show why ToM evaluation must connect perspective inference to independently verified decisions and task consequences.
Knowing when they are reliable is essential for language models to support human-AI collaboration. Yet interaction feedback is not always complete or correct, and may mislead models' understanding of their own capabilities. We introduce LADDER to study how language models assess themselves under unreliable feedback. Our findings show that models can accurately report supplied performance information while still struggling to form trustworthy self-assessments from interaction experience. Apparently accurate assessments can obscure capability differences and do not necessarily predict future task performance. These findings suggest that trustworthy self-assessment should be evaluated against observed task performance and its predictive value. Our study provides empirical evidence of the limits of model self-knowledge and a basis for evaluating capability awareness in reliable human-AI collaboration.
Zhihao Ding, Jinming Li, Ze Lu, Jieming Shi The Hong Kong Polytechnic University
Ensuring the safety of LLM-generated content is essential for real-world deployment. Most existing guardrail models formulate moderation as a fixed binary classification task, implicitly assuming a fixed definition of harmfulness. In practice, enforcement strictness—how conservatively harmfulness is defined and enforced—varies across platforms and evolves over time, making binary moderators brittle under shifting requirements. In this paper, we introduce FlexBench, a strictness-adaptive LLM moderation benchmark that enables controlled evaluation under multiple strictness regimes. Experiments on FlexBench reveal substantial cross-strictness inconsistency in existing moderators: models that perform well under one regime can degrade substantially under others, limiting their practical usability. To address this, we propose FlexGuard, an LLM-based moderator that outputs a calibrated continuous risk score reflecting risk severity and supports strictness-specific decisions via thresholding. We train FlexGuard via risk-alignment optimization to improve score–severity consistency and provide practical threshold selection strategies to adapt to target strictness at deployment. Experiments on FlexBench and public benchmarks demonstrate that FlexGuard achieves higher moderation accuracy and substantially improved robustness under varying strictness.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表