Lang3DSeg introduces a point‑transformer backbone for open‑vocabulary, annotation‑free 3D LiDAR segmentation, trained from scratch without geometric pre‑training. It tackles noise from 2D‑to‑3D label projections by applying a class‑priority rule and truncating projected instances at depth gaps, thereby correcting depth‑ambiguity errors. The method achieves state‑of‑the‑art results on nuScenes (52.8 % mIoU) and SemanticKITTI (41.4 % mIoU) while operating in real‑time on a single LiDAR sweep.
By Cigdem Kokenoz, Amir Salarpour, Alkim Domeke, Christopher Salas, Pedram MohajerAnsari, Long Cheng, Mert D. Pes\'e, Bing Li
The paper reviews end‑to‑end autonomous driving (E2E‑AD) training, framing it as a Data‑Strategy‑Platform system. It surveys recent advances in data pipelines, learning paradigms, and training infrastructures, and discusses how these layers interact to influence model performance, robustness, and deployability. The authors highlight current limitations and propose a future vision that prioritizes data value, foundation‑driven generalization, and integrated training‑testing loops for more robust, scalable, and trustworthy autonomous driving systems.
By Chengkai Xu, Yiming Cui, Jiaqi Liu, Yicheng Guo, Cheng Qin, Geyuan Zhang, Xinwei Dong, Shiyu Fang, Peng Hang, Jian Sun
The paper introduces GTA‑VLA, an interactive Vision‑Language‑Action framework that lets users guide robot policies with explicit visual cues such as affordance points, boxes, and traces. Unlike traditional direct sense‑to‑act models, GTA‑VLA incorporates a spatial‑visual Chain‑of‑Thought that blends human guidance with internal task planning, and couples this reasoning module with a lightweight reactive action head for efficient execution. Experiments on the SimplerEnv WidowX benchmark show a state‑of‑the‑art 81.2 % success rate, and the framework significantly improves task success under out‑of‑domain visual shifts and spatial ambiguities, demonstrating the benefit of interactive reasoning for failure recovery in embodied control.
By Yiran Ling, Qing Lian, Jinghang Li, Qing Jiang, Tianming Zhang, Xiaoke Jiang, Chuanxiu Liu, Jie Liu, Lei Zhang
Build2SPARQL is a large-scale benchmark dataset for translating natural-language questions into SPARQL queries over building knowledge graphs. The dataset is generated by a KG‑grounded pipeline that produces 6,136 executable SPARQL queries and 30,680 corresponding natural-language questions across six query-pattern families and five vocabulary registers, covering 201 building KGs. Human validation shows high semantic fidelity, naturalness, and operational plausibility, and retrieval‑augmented evaluation demonstrates significant accuracy gains for open‑weight language models.
By Wooyoung Jung
Praxa is an evidence‑bound harness for governed AI agent execution that explicitly represents states such as proposal, authority, dispatch, verified external effect, and promotion through deterministic admission, brokered execution, external read‑back, reconciliation, and reviewed promotion. The authors report four evidence lanes: a repository‑local audit passing all unit and Workerd tests; a pilot on 12 curated tasks where both baseline and reliability‑layer arms passed 17 of 36 trials; a coordination‑proxy comparison where both baseline and a source‑authored candidate completed all 180 trials with equal accuracy but the candidate used fewer tokens and steps; and deployed source/configuration evidence showing bounded reflection, recall accounting, memory compilation, and tool‑health paths. None of the evidence demonstrates superiority in security, safety, or user benefit.
whyItMatters:"Praxa provides a testable architecture that makes authority‑to‑effect transitions explicit, offering a framework for verifying AI agent behavior, though current evidence does not prove improved security or performance."
By Stefan G. Creadore
JusticeAxis is a new benchmark that formalizes legal judgment as a reference‑anchored task, requiring a single decision tied simultaneously to the statute and the surrounding circumstances. It comprises 256 real‑world criminal cases from 18 countries, each with audio, image, and text evidence, and three lawyer‑written judgments per case: the recorded judgment and two failure judgments. The authors also introduce JusticeAgent, a modular system that first establishes facts and then applies the law, with skills distilled from execution trajectories and validated under Bayesian credible bounds. Experiments reveal that open‑weight backbones tend to drift toward unsupported grounds while frontier models default to statutory interpretations, and that JusticeAgent can effectively plug into commercial systems.
By Zhengkai Tu, Mingda Zhang, Zijia Wang, Xiaoying Tang, Jimmy Huang
JevSpawn is a new compositional policy that links natural language task specifications to finite probabilistic exploration, enabling LLM agents to generate actions more efficiently. It uses parallel action spawning, feedback‑driven branch selection, representation revision, and recovery from retained alternatives to adapt actions during interaction. Evaluations on eight benchmark tasks show that JevSpawn outperforms seven agent baselines and a TypeSafe Jev variant, improving task performance and speeding navigation.
By Haoyang Su, Weiran Huang
The paper investigates how coordination among AI agents serving different users degrades performance compared to a single coordinating agent. Across five advanced models and 77 scenarios in four shared-resource environments—API key budgets, clinic calendars, personal assistant bookings, and merge queues—the study finds that multi‑agent teams consistently underperform, sometimes collapsing entirely, and that even with communication channels coordination overhead remains significant. The authors identify specific failure modes such as stalling, action overriding, and claim fabrication, and propose environment‑specific mitigations like team leads and procedural instructions, while releasing the MAMUBench benchmark for future research.
By Sahan Paliskara, Nattaput Namchittai, Andrew Lampinen
CompMat-Bench is a new benchmark comprising 94 tasks drawn from recent computational materials science studies, designed to evaluate AI agents on realistic research steps without requiring costly simulations during testing. The benchmark pre‑reproduces inputs and outputs to provide ground truth, allowing agents to be graded with fixed rules rather than an LLM judge. It supports both single tasks and multi‑step workflows, with varying levels of methodological guidance, and shows that while agents can achieve high pass rates on individual tasks, performance drops in longer workflows or with reduced guidance, often due to scientific rather than software errors.
By Chenmu Zhang, Levi Felix, Jun-Jie Zhang, Xingfu Li, Xuelian Jiang, Tao Jiang, Subhendu Mishra, Xixi Qin, Boris Yakobson
Sapien is a policy engine that enforces stateful contextual policies for autonomous AI agents, specifying allowed tool‑call sequences with an extended regular expression that includes stateful predicates, deferred policy generation, and scoped semantic checks. The system maintains performance close to an unconstrained agent while significantly reducing malicious actions, ruling out 93‑95% of attacks on AgentDojo and 62‑85% on Toolathlon, outperforming traditional tool allowlists on long‑horizon tasks.
By Corinn Tiffany, Wen Zhang, Eugene Bagdasarian, Lillian Tsai
Kepler is an open‑source harness that represents hypotheses as executable world models and validates them through retrospective transition checks and conditional prediction checks. In the ARC‑AGI‑3 benchmark, a frozen Claude Opus 5 configuration achieved a perfect 100.00 RHAE on all 25 public games without per‑game model selection or score‑conditioned reruns, and matched or outperformed median‑human action counts on 181 of 183 levels. The study also identified three evaluation failures and highlighted that public‑set score alone has limited discriminative value, advocating for first‑attempt, cost‑conditioned, and verification‑aware reporting.
whyItMatters":"The results demonstrate that a purely score‑based evaluation can be misleading, underscoring the need for more rigorous, cost‑aware, and verification‑aware metrics in AI benchmark assessments."
By Wensen Wu
PG-SFT: Balancing Capability Acquisition and Retention in Offline Agent Fine-Tuning explores how to maintain a model’s existing abilities while teaching it new ones through supervised fine‑tuning on offline agent trajectories. The authors compare standard SFT, KL‑penalty, and update‑magnitude constraints, finding that these methods still degrade non‑target capabilities. They introduce Privilege‑Guided SFT (PG‑SFT), which uses turn‑level information gain to modulate supervision strength, achieving a better trade‑off between acquiring new skills and preserving existing ones, though with a slight drop in target‑task performance.
By Ronghua Li, Zi Liang, Zhishan Li, Shinan Liu
RISED introduces a framework that uses rubric-based textual feedback to improve training of a single large language model (LLM) agent across multiple interactive environments. By having an LLM judge tag rollouts with a shared rubric vocabulary, the system guides both online data selection and policy supervision, enabling richer cross‑environment relationships and within‑group reward contrast. Experiments show that RISED achieves the highest mean pass rate and ranks first or second in every individual environment, with rubric analysis revealing behavioural changes behind these gains.
By Jingtan Wang, Sirajul Salekin, Young mok Jung, Javier Movellan, Bryan Kian Hsiang Low, Manjot Bilkhu
The paper introduces the concept of empty commitments, where an agent promises an action that its current tools or runtime cannot execute. It defines empty commitments using commitment semantics, identifies three failure types, and proposes an anchoring condition for realistic tool promises. The authors also outline a measurement protocol that tests follow‑up requests across five setups, each adding a different persistence affordance, with the environment either implicit or explicitly stated.
By Jiaqi Tang, Lan Wei, Bingyu Shen, Boyang Li
The paper introduces an action‑conditioned Network World Model that learns how a network’s diffusion dynamics evolve under interventions over time. This model can quickly predict the outcomes of actions, enabling a coding agent to design and refine algorithms that select actions to maximize expected performance on complex network tasks. Experiments on eight network tasks and five diffusion models show that the resulting algorithms match or surpass the best existing baselines in 138 of 141 settings while achieving up to 14.5× faster rollouts than traditional Monte Carlo simulation.
By Rishab Alagharu, Hongji Pu, Zeeshan Memon, Xinyuan Song, Yuntong Hu, Liang Zhao
AbsorbEvo is an agentic framework that autonomously designs microwave absorbers by translating natural‑language performance goals into full‑wave‑simulated designs. It combines language reasoning, physics‑based prediction, and historical feedback to guide a candidate evolution strategy, using a low‑cost predictive model to rank designs before simulation. In tests on AbsorbBench‑36, AbsorbEvo achieved a 79.17% task success rate, outperforming generic agents and random search.
By Zhicheng Feng, Yubo Zhao, Xuefeng Yao
The paper introduces Dependency‑Aware Reward Shaping (DARS), a method that assigns step‑level credit in reinforcement learning by modeling task progress as a graph of predicates with prerequisite relations. Annotators mark each step’s effect on predicates, and DARS discounts verified predicates based on distance from broken prerequisites while preserving independent ones, converting these annotations into signed per‑step rewards. Experiments on five task families with models ranging from 1.5B to 8B show that DARS improves success rates by up to 10 points over GiGPO, boosts WebShop and Search‑R1 QA scores, complements AEPO on AIME24/25, and outperforms OmniOPD in tool‑free reasoning, with ablations confirming the contribution of step‑level credit, dependency attenuation, and graph topology.
By Ziyi Chen, Yan Zhang, Jianhui Wei, Daoan Zhang, Zuozhu Liu
The study investigates how reputation, strategy, and emotional signals influence cooperation in generative AI models using the iterated prisoner's dilemma. Non‑reasoning models (Claude 3.5, Gemini 2.0 Flash, GPT‑4o) showed cooperation shaped by all three factors, while reasoning models (Claude 4.6, Gemini 3, GPT‑5.2) relied more on strategy and reputation, displayed reduced emotional influence, and exhibited varied end‑game behaviors. These results highlight the growing sophistication and heterogeneity of AI social behavior, suggesting the need for standardized cooperation benchmarks.
By Celso de Melo, Zishan Feng, James Hale, Kazunori Terada, Giorgio Coricelli, Jonathan Gratch
The paper introduces Revision‑Aware Independent Agent Graphs (RIAG) to address dynamic task routing, where an event stream continually revises task bindings and a system must select the correct document version at query time. By repurposing six benchmarks into over 31,000 dynamic episodes, the authors demonstrate that RIAG balances recomputation and reuse, achieving 54.24 % joint routing‑and‑answer accuracy with only 0.62 calls per query—substantially better than the strongest baseline. The study highlights the trade‑off between stale conclusions and wasted work in dynamic reasoning settings.
By Yan Luo, Selim-Antoine Lali, Jeremy Moebel, Iliass Khoutaibi, Ahmadou Aidara, Mengyu Wang
DAYJOB is a new benchmark comprising 130 long‑horizon professional tasks created by experts in healthcare (50 tasks) and finance (80 tasks). Each task is packaged as a containerized Harbor environment and evaluated against a detailed binary rubric, requiring agents to meet every criterion to pass. In tests across 30 model configurations, the best model (Claude Opus 5.5) achieves about 24% success in both domains, while the median model scores below 3%.
By Stephanie Finley, Liudas Panavas, Thomas Mikkelson, Cam Hinton, Stacey Ganss, Bradley Monton, Emily Kendall, Michelle Spradlin, Lydia Bye, Michael O'Brien, Lauren Ylvisaker, Derek Ray, Suhaas Garre, Sushant Mehta, Edwin Chen