Yijun Zhou, Minghao Shao, Ke Zhang, Renkai Yuan, Hongbo Yuan, Bin Hu, Bozhao Qian, Hongxia Shen 等 23 位作者
Game development stresses coding agents in ways general software engineering does not: correctness depends on how engine APIs, serialized scene state, and runtime behavior interact, not on source code alone. Existing benchmarks mainly target platforms such as Godot and heavily rely on LLM judges or coarse task-level outcomes. We present UnityBench-Mini3D, a benchmark of 77 project-level Unity tasks organized around seven expert-defined skill categories rather than tutorials or genres, with three difficulty tiers. Each task pairs a complete Unity project with natural-language requirements and is evaluated using 1,178 held-out, human-authored milestone tests covering scene structure, component configuration, and runtime behavior. This in-engine evaluation provides deterministic partial credit, localizes failures, and distinguishes partial implementations from complete solutions without relying on subjective judges. Across 21 models and 9 coding harnesses, the average pass rate reaches 80.9% on easy tasks but falls to 35.2% on hard tasks, with just 57.7% overall. UnityBench-Mini3D provides a reproducible, diagnostic testbed distinguishing systems across difficulty levels.
Junxing Hu, Zicheng Zhang, Zhiqian Zhang, Ai Han, Yifeng Zhang, Pengzhang Liu, Ling Li
Reasoning models trained with Group Relative Policy Optimization (GRPO) now produce chains of thought thousands of tokens long, and recent studies reveal that such training is prone to policy collapse. Entropy regularization is the standard remedy, but it works by pushing probability mass away from the few tokens that carry it and toward the vast low-probability tail of the vocabulary, and it applies this force once for every token of the reasoning trace. This puts the bonus in direct conflict with the entropy decay that convergence requires, and makes its coefficient difficult to set in the long-trace regime. To resolve this tension, we propose Info-GRPO, an information-theoretic framework that cultivates correlation between the policy and a latent prior. Rather than raising entropy everywhere, Info-GRPO builds on the mutual information between a latent variable and the trajectory it induces: augmenting prompts with latent seeds during training lets the model explore diverse policies correlated with the prior, while the seed-conditioned entropy is guided toward convergence. Diversity is therefore carried across whole reasoning traces rather than by per-token randomness, and exploration no longer fights entropy reduction. Experiments show that Info-GRPO outperforms vanilla GRPO and entropy-regularized GRPO across diverse reasoning benchmarks. On AIME24, it improves $Avg@8$ over GRPO by 3.75%, 1.66%, 4.16%, and 4.17% on four backbones, and it reaches the same majority-vote accuracy with $8\times$ fewer samples. Code and models will be publicly released.
Yixue Zhang, Kun Wu, Zhi Gao, Zhen Zhao, Pei Ren, Zhiyuan Xu, Fei Liao, Xinhua Wang 等 15 位作者
The pursuit of general-purpose robotic manipulation is hindered by the scarcity of diverse, real-world interaction data. Unlike large-scale web vision-language data, robotic data collection incurs prohibitive physical costs. The key challenge is thus not merely to collect more robot data, but to obtain informative training data that broadens behavioral coverage across objects, scenes, and skills. Yet automatically curating task proposals to maximize the utility of collected data remains under-explored. We introduce RoboGene, an agentic framework for generating diverse, informative manipulation tasks across single-arm, dual-arm, and mobile robots. RoboGene combines (1) diversity-driven sampling for broad task coverage, (2) agentic self-reflection for task-quality assessment, and (3) human-verified feedback memory for execution-grounded validation and continual refinement. Through quantitative analysis and large-scale real-world experiments, we collect 18k trajectories and introduce metrics for task feasibility and diversity. RoboGene significantly outperforms frontier foundation models, including GPT-4o and Gemini 2.5 Pro. VLA models trained on RoboGene-generated data achieve higher success rates and stronger generalization, highlighting the value of high-quality task generation. Project page: https://robogene-boost-vla.github.io/. Datasets: Hugging Face (https://huggingface.co/datasets/X-Humanoid/RoboGene) and ModelScope (https://modelscope.cn/datasets/X-Humanoid/RoboGene_tasks).
Zhanwei Wu, Jianhao Su, Brian Zhang, ShengTing Huang, Weidong Feng, Gang Liu, Zhiguo Li, Wei Huang 等 14 位作者
Deploying and optimizing a PyTorch inference project on a dedicated accelerator, especially an on-device NPU, requires more than using a framework device API as with CUDA. The project must be converted into hardware-runtime artifacts and adapted to the target toolchain and application interface. Differences in model formats, quantization flows, operator support, and toolchain interfaces create cross-layer problems spanning PyTorch/ONNX models, toolchains, runtimes, and device execution. LLM Agents can invoke tools to address such problems, but may drift because of insufficient domain knowledge, large configuration and repair spaces, or poor path selection. We propose AIPC (AI Porting & Conversion), an Agent method for deploying PyTorch inference projects. AIPC combines structured Skills with toolchain, runtime, and profiling evidence to drive cross-layer diagnosis and repair, organizing model adaptation, artifact building, on-target execution, accuracy repair, and performance optimization into one deployment flow. We evaluate AIPC on the on-device NPU path of Qualcomm QAIRT across vision, speech, LLM, and diffusion models. In two optimization cases, 6/7 Skill-arm Whisper-Small runs meet predefined accuracy and latency criteria versus 0/6 control runs, while YOLO-Worldv2 achieves a 2.791x end-to-end speedup versus 1.599x for the control. Under a pre-fixed deployment contract, Skill, tool permissions, and check conditions, some runs achieve zero in-run human intervention (0-HITL). We also examine Qwen3.8-27B as a locally deployable model for driving an Agent workflow.
Chenxi Liu, Yongqiang Chen, Tongliang Liu, Mingming Gong, James Cheng, Bo Han, Kun Zhang
Revealing hidden causal variables alongside the underlying causal mechanisms is essential to the development of science. Despite the progress in the past decades, existing practice in causal discovery (CD) heavily relies on high-quality measured variables, which are usually given by human experts. In fact, the lack of well-defined high-level variables behind unstructured data has been a longstanding roadblock to a broader real-world application of CD. This procedure can naturally benefit from an automated process that can suggest potential hidden variables in the system. Interestingly, Large language models (LLMs) are trained on massive observations of the world and have demonstrated great capability in processing unstructured data. To leverage the power of LLMs, we develop a new framework termed Causal representatiOn AssistanT (COAT) that incorporates the rich world knowledge of LLMs to propose useful measured variables for CD with respect to high-value target variables on their paired unstructured data. Instead of directly inferring causality with LLMs, COAT constructs feedback from intermediate CD results to LLMs to refine the proposed variables. Given the target variable and the paired unstructured data, we first develop COAT-MB that leverages the predictivity of the proposed variables to iteratively uncover the Markov Blanket of the target variable. Built upon COAT-MB, COAT-PAG further extends to uncover a more complete causal graph, i.e., Partial Ancestral Graph, by iterating over the target variables and actively seeking new high-level variables. Moreover, the reliable CD capabilities of COAT also extend the debiased causal inference to unstructured data by discovering an adjustment set. We establish theoretical guarantees for the CD results and verify their efficiency and reliability across realistic benchmarks and real-world case studies.
Chenyang Fan, Robert D. Sanders, Elie Matar, Yixuan Yuan, Xiang Li, Sharon L. Naismith, Fernando Calamante, Jinglei Lv
Diffusion MRI (dMRI) provides information about white matter microstructure that can support brain age prediction. However, many existing prediction models use diffusion tensor imaging (DTI) scalar metrics that omit orientation information. Deep learning is increasingly used for brain age prediction, but conventional CNNs do not explicitly respect the geometric structure of orientation-dependent features. We present an interpretable brain age prediction framework that combines white matter fiber orientation distributions (WM-FODs) with an SE(3) (Special Euclidean group in 3 dimensions) equivariant CNN. The network operates directly on spherical harmonic (SH) coefficients of the FOD, preserving their transformation properties under rotation. Evaluation in 692 participants aged 36-90 years from the Human Connectome Project Aging dataset showed that the combined framework outperformed the evaluated DTI-based models and conventional CNNs. Averaging predictions from five cross-validation models yielded a mean absolute error of 3.54 years and Pearson's correlation of 0.958. Our results indicated that the performance advantage depended on combining WM-FODs with equivariant processing. Prediction accuracy was highest at a maximum SH order of Lmax = 2 and declined at higher orders. Tract-level occlusion, where each segmented tract was masked in turn, revealed age-dependent patterns of model sensitivity, with importance peaks in frontal projection pathways at 36-44 years, the fornix at 45-54 years, the callosal isthmus at 55-64 years, the middle cerebellar peduncle at 65-74 years, and the callosal genu at 75-90 years. These findings support rotation-equivariant processing of WM-FODs for accurate brain age prediction with anatomically interpretable model sensitivity.
Guided depth super-resolution (GDSR) typically extracts RGB guidance features through convolutional hierarchies, inheriting their fine-to-coarse bias. Thus, low-level spatial cues surface in early layers, leaving the deeper layers to suppress those that do not correspond to true depth boundaries, which risks artifacts and blurred edges. The same fine-to-coarse bias persists in semantics-based methods that consume low-level tokens early and global tokens late. We present WAVE, which introduces a multi-level discrete wavelet transform (ML-DWT) as an explicit and interpretable feature-control mechanism, enabling a coarse-to-fine reconstruction by consuming sub-bands and semantic tokens in reverse of their generation order. WAVE further exploits these sub-bands to treat high- and low-frequency content separately, isolating the misleading RGB color and texture cues before any feature is learned so that their admission can be controlled explicitly, offering an intuitive alternative to the suppression learned implicitly by an opaque network. WAVE separates structure and detail reconstruction into dedicated modules that: i) model interactions within and across wavelet sub-bands, depth features, and semantic priors, ii) apply semantic gating to the high-frequency bands, and iii) fuse modalities through an invertible coupling mechanism that prevents collapse onto a single modality. Extensive experiments across multiple benchmarks demonstrate that WAVE matches or outperforms existing methods, with the largest gains at high upsampling factors, where low-resolution depth contains the least structure.
Edric Tam, Qingyang Liu, Barbara Engelhardt, Ray Bai
Broad classes of deep generative models (DGMs) operate by learning to map a simple source distribution, such as a standard Gaussian, to a complex high-dimensional target distribution. Despite their success, DGMs often empirically struggle to accurately represent distributions with heavy tails or well-separated modes. In this work, we show that these difficulties reflect intrinsic properties of the underlying transport problem rather than particular neural network architectures or training procedures. We establish quantitative lower bounds on the gradients and Jacobians of pushforward mappings that transport log-concave source distributions to such targets. For heavy-tailed target distributions, the required gradient magnitude grows as the target develops increasingly large moments. For multimodal target distributions, the low-probability regions between well-separated modes similarly force large gradients in the transport map. We discuss how such large derivatives relate to trade-offs between expressivity and numerical stability. Our results hold under substantially weaker regularity conditions than previous analyses and remain stable under perturbations of the target distribution. Finally, we show that these obstructions can be mitigated when the goal is approximation of a given density rather than generative modeling. Specifically, we propose simple classes of bijective reparametrizations that can alleviate the difficult geometric structure and reduce the gradient burden placed on the learned generative map. Numerical experiments support our theoretical analysis.
George Nigmatulin, Valentin Mohl, Sascha Frey, Kang Li, Aramis Marti-Shahandeh, Mihai Cucuringu, Anisoara Calinescu, Jakob Foerster 等 9 位作者
Realistic market impact is essential for generative limit order book (LOB) models to serve as world models for trading agents. If a simulator misprices this mechanism, agents learn policies that are optimal in the simulator but wrong in the real market. Existing metrics assess generated streams purely through distributional comparison with historical data, leaving counterfactual response to an agent's own orders largely unexplored. Here, we show that market impact emerges naturally in neural LOB generators trained solely for next-token prediction. We inject execution-style metaorders into the generated message flow of five autoregressive models across five NASDAQ stocks and measure the response using estimators from the empirical impact literature, validated on baselines whose physics is known in advance. Our benchmark evaluates any message-level generator in three phases: free generation (reading order-flow persistence against price diffusivity); metaorder injection (yielding the impact law's exponent and amplitude); and a post-execution window (measuring relaxation). In addition, comparing each visible injection to a blind, simulator-only control isolates the model's learned response from standard book mechanics. For every model on every stock, we measure how the impact responds as a function of cumulative metaorder participation. Interestingly, we observe emergent impact in generative models of market data, with exponents in the range of 0.63 to 1.21, slightly above the traditional 0.5 of the square-root law of impact. In the family of models tested, we are so far unable to observe impact decay during relaxation. We tentatively attribute this to the still-short context windows of the models. We release the benchmark as open source to validate future generative models for market data.
Tony Feng, Trieu H. Trinh, Garrett Bingham, Dawsen Hwang, Yuri Chervonyi, Junehyuk Jung, Joonkyung Lee, Carlo Pagano 等 28 位作者
Recent advances in foundational models have yielded reasoning systems capable of achieving a gold-medal standard at the International Mathematical Olympiad. The transition from competition-level problem-solving to professional research, however, requires navigating vast literature and constructing long-horizon proofs. In this work, we introduce Aletheia, a math research agent that iteratively generates, verifies, and revises solutions end-to-end in natural language. Specifically, Aletheia is powered by an advanced version of Gemini Deep Think for challenging reasoning problems, a novel inference-time scaling law that extends beyond Olympiad-level problems, and intensive tool use to navigate the complexities of mathematical research. We demonstrate the capability of Aletheia from Olympiad problems to PhD-level exercises and most notably, through several distinct milestones in AI-assisted mathematics research: (a) a research paper (Feng26) generated by AI without any human intervention in calculating certain structure constants in arithmetic geometry called eigenweights; (b) a research paper (LeeSeo26) demonstrating human-AI collaboration in proving bounds on systems of interacting particles called independent sets; and (c) an extensive semi-autonomous evaluation (Feng et al., 2026a) of 700 open problems on Bloom's Erdos Conjectures database, including autonomous solutions to four open questions. In order to help the public better understand the developments pertaining to AI and mathematics, we suggest quantifying standard levels of autonomy and novelty of AI-assisted results, as well as propose a novel concept of human-AI interaction cards for transparency. We conclude with reflections on human-AI collaboration in mathematics and share all prompts as well as model outputs at https://github.com/google-deepmind/superhuman/tree/main/aletheia.
Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang 等 10 位作者
Test-time scaling is a promising new approach to language modeling that uses extra test-time compute to improve performance. Recently, OpenAI's o1 model showed this capability but did not publicly share its methodology, leading to many replication efforts. We seek the simplest approach to achieve test-time scaling and strong reasoning performance. First, we curate a small dataset s1K of 1,000 questions paired with reasoning traces relying on three criteria we validate through ablations: difficulty, diversity, and quality. Second, we develop budget forcing to control test-time compute by forcefully terminating the model's thinking process or lengthening it by appending "Wait" multiple times to the model's generation when it tries to end. This can lead the model to double-check its answer, often fixing incorrect reasoning steps. After supervised finetuning the Qwen2.5-32B-Instruct language model on s1K and equipping it with budget forcing, our model s1-32B exceeds o1-preview on competition math questions by up to 27% (MATH and AIME24). Further, scaling s1-32B with budget forcing allows extrapolating beyond its performance without test-time intervention: from 50% to 57% on AIME24. Our model, data, and code are open-source at https://github.com/simplescaling/s1
Hammer: Robust Function-Calling for On-Device Language Models via Function Masking
Qiqiang Lin, Muning Wen, Qiuying Peng, Guanyu Nie, Junwei Liao, Jun Wang, Xiaoyun Mo, Jiamu Zhou 等 12 位作者
AI 速览：提出 Hammer 系列端侧函数调用基础模型，通过增强数据集和函数掩码技术提升对无关函数的敏感性和跨基准泛化能力。
Large language models have demonstrated impressive value in performing as autonomous agents when equipped with external tools and API calls. Nonetheless, effectively harnessing their potential for executing complex tasks crucially relies on enhancements in their function calling capabilities. This paper identifies a critical gap in existing function calling models, where performance varies significantly across benchmarks, often due to being misled by specific naming conventions. To address such an issue, we introduce Hammer, a novel family of foundation models specifically engineered for on-device function calling. Hammer employs an augmented dataset that enhances models' sensitivity to irrelevant functions and incorporates function masking techniques to minimize misleading. Our empirical evaluations reveal that Hammer not only outperforms larger models but also demonstrates robust generalization across diverse benchmarks, achieving sota results. Our open source contributions include a specialized dataset for irrelevance detection, a tuning framework for enhanced generalization, and the Hammer models, establishing a new standard for function calling performance.
Peter Kairouz, H. Brendan McMahan, Brendan Avent, Aurélien Bellet, Mehdi Bennis, Arjun Nitin Bhagoji, Kallista Bonawitz, Zachary Charles 等 59 位作者
AI 速览：一篇综述性论文，梳理联邦学习的研究进展并系统列出开放问题与挑战。
Federated learning (FL) is a machine learning setting where many clients (e.g. mobile devices or whole organizations) collaboratively train a model under the orchestration of a central server (e.g. service provider), while keeping the training data decentralized. FL embodies the principles of focused data collection and minimization, and can mitigate many of the systemic privacy risks and costs resulting from traditional, centralized machine learning and data science approaches. Motivated by the explosive growth in FL research, this paper discusses recent advances and presents an extensive collection of open problems and challenges.