Artificial Intelligence · Cross-listed from cs.SE
Lossless Tensor Compression as Program Synthesis
Jieke Shi, Junda He, Wenjia Jiang et al. (+11 more)
Model checkpoints are growing in both number and size, which makes archival, transfer, and deployment increasingly costly. General-purpose compressors can reduce storage requirements but ignore tensor structure, whereas existing tensor-specific compressors rely on fixed and format-specific pipelines. We present Brevis, which formulates lossless tensor compression as program synthesis. We design a typed domain-specific
language (DSL) that captures recurring tensor structures, such as repeated regions and floating-point fields, through a set of reversible operators. Given a tensor, Brevis synthesizes a self-contained DSL program that reconstructs it bit-exactly. A checkpoint-specific production prior, learned from a small representative sample of tensors, guides a bounded A* search to synthesize compact programs, which can later be executed directly for bit-exact decompression. On 10 public checkpoints spanning
language, audio, and image generation models, Brevis reduces 2.13 TB of checkpoint data to 1.41 TB, a 33.93% storage reduction. It produces archives up to 30.87% smaller than those of four general-purpose compressors, including zstd and gzip, and smaller archives than the tensor-specific compressors ZipNN and DFloat11. Under a practical concurrency configuration, Brevis achieves 3.60 GB/s compression and 6.61 GB/s decompression while preserving every source byte.
Read the full abstract ▾
All authors Jieke Shi, Junda He, Wenjia Jiang, Weifeng Sun, Shidong Pan, Zhensu Sun, Chengran Yang, Peixin Zhang, Yifan Jia, Zhou Yang, Thong Hoang, Xiwei Xu, Zhenchang Xing, David Lo
Artificial Intelligence · Cross-listed from cs.LG
RamanPFN: learning from Raman spectral structure with a tabular foundation model
Xingyu Pan, Huan Wang, Jinjia Guo et al. (+3 more)
Raman spectroscopy enables label-free molecular characterization across materials science, analytical chemistry, biomedicine, and industrial process monitoring. However, machine learning for high-dimensional spectroscopy remains constrained by limited labelled data and a mismatch between the physical organization of spectra and feature-agnostic models. Channel coverage alone does not ensure that related bands share a common inference context. Here we present RamanPFN, a general-purpose spectral foundation framework that enables unified in-context inference through physics-guided spectral learning. It captures full-spectrum compositional covariation via Global Compositional Unmixing (GCU), which decomposes distributed, multi-band mixture signatures into shared non-negative latent bases. Simultaneously, it resolves local vibrational structure through Local Vibrational Subspace Encoding (LVSE), which preserves fine-grained peak morphology, intensity fluctuations, and peak shifts within contiguous spectral neighborhoods. Extensive evaluation across 74 diverse public Raman
datasets covered 129 regression targets and was further extended to 21 classification tasks. RamanPFN achieved state-of-the-art performance across all reported aggregate metrics against 28 independently reproduced methods spanning chemometrics, spectral neural networks, deep tabular learners and tabular foundation models. RamanPFN establishes a physics-guided paradigm for scientific spectroscopy, enabling data-efficient predictive learning across diverse chemical systems.
Read the full abstract ▾
All authors Xingyu Pan, Huan Wang, Jinjia Guo, Zhenlin Zhao, Siming Dong, Jixi Lu
Artificial Intelligence
Auditing Data Provenance in LLM Fine-tuning via Intrinsic Distributional Fingerprints
Zirui Huang, Yunlong Mao, Wei Tong et al. (+3 more)
The proliferation of customized Large
Language Models (LLMs) poses critical risks of Data Intellectual Property (Data IP) infringement via unauthorized fine-tuning on proprietary data. Existing audit techniques are limited, as they require intervention during data preparation or training and remain fragile under malicious obfuscations such as data paraphrasing and knowledge
distillation. We propose Distribution Provenance Audit (DPA), a post-hoc framework for
auditing data IP infringement in LLM fine-tuning under black-box and malicious settings. DPA is grounded in a critical insight: regardless of fine-tuning tactics to evade provenance, the practical necessity of maintaining utility constrains the model to preserve the fundamental intersection of semantic substance and lexical form. Accordingly, DPA captures this persistent lexical-semantic intersection as intrinsic distributional fingerprints. The framework formulates the audit as a statistical hypothesis test, effectively quantifying these fingerprints via unbiased output sampling to reliably reject the null hypothesis of non-usage. Extensive experiments on medical and legal fine-tuning tasks show that DPA consistently outperforms existing baselines, remaining robust against
adversarial trainers employing paraphrasing and knowledge
distillation. We further highlight a fundamental dual-use tension: the same high-fidelity distributional fingerprints enabling reliable
auditing may also facilitate
privacy attacks.
Read the full abstract ▾
All authors Zirui Huang, Yunlong Mao, Wei Tong, Tingting Wu, Xin Ge, Sheng Zhong
Artificial Intelligence · Cross-listed from cs.CV
PhyCheck: Fine-Grained Evidence-Grounded Dataset for Physical Law Understanding in Video-LLMs
Zhongjie Ba, Shengwang Xu, Peng Cheng et al. (+4 more)
Embodied intelligence and world models require
video understanding systems to go beyond recognizing objects and actions and develop an understanding of physical regularities. However, despite their strong performance on general
video understanding tasks, current
video-
language models still struggle to reliably determine whether an observed event conforms to specific physical laws. Existing
benchmarks primarily assess the physical quality of generated
videos, providing limited support for systematically evaluating and improving the physical-law understanding of
Video Large
Language Models (VideoLLMs). To address this gap, we introduce PhyCheck, a
video question answering
dataset organized at two complementary levels of granularity. The coarse-grained subset asks models to determine whether the phenomenon shown in a
video conforms to or violates physical laws, while the fine-grained subset further examines whether models can capture physical details responsible for the violation or compliance. We use these subsets as structured supervision to improve physical understanding. In addition, the
dataset contains a diagnostic subset with external causal context that reveal hidden factors affecting physical plausibility, assessing whether models can recalibrate their judgments accordingly. Experiments with Fine-tune Qwen2.5-VL show that training with the proposed data substantially improves the understanding of physical-consistency, while evaluations in the diagnostic subset reveal that current models still have difficulty incorporating additional causal conditions into their decisions. These findings highlight the gap between recognizing surface-level inconsistencies and understanding underlying physical mechanisms, and provide a foundation for evaluating and improving physical understanding in
Video-LLMs.
Read the full abstract ▾
All authors Zhongjie Ba, Shengwang Xu, Peng Cheng, Jinyang Zou, Ting Yu, Zhibo Wang, Zhan Qin
Artificial Intelligence
Beyond the Mean: Multi-Moment Policy Optimization for LLM Reasoning
Yijun Zhang, Yule Xie, Jiaxin Ding et al. (+4 more)
Reinforcement learning has become a central paradigm for improving the
reasoning capabilities of large
language models. Existing methods generally aim to reduce the failure probabilities induced across problems. In this paper, we introduce a moment-based perspective on policy optimization for LLM
reasoning by treating the failure probability of a randomly sampled problem as a random variable and characterizing optimization objectives through its moments. Under this perspective, many existing methods optimize only a single moment of the failure-probability distribution, leaving its broader distributional structure largely uncharacterized. We propose Multi-Moment Policy Optimization (MMPO), a novel policy optimization framework that jointly minimizes multiple moments of the failure-probability distribution. MMPO admits a direct operational interpretation as minimizing the expected truncated time required to obtain the first successful response. Beyond MMPO, we further develop a general moment-transformation framework that systematically induces different moment profiles and provides a unified view of a broader family of policy optimization objectives. Experiments across five mathematical
reasoning benchmarks and models of different scales demonstrate that MMPO consistently outperforms strong baselines. We hope this moment-based perspective offers new insights into the design of policy optimization objectives for LLM
reasoning.
Read the full abstract ▾
All authors Yijun Zhang, Yule Xie, Jiaxin Ding, Xin Ding, Fan Xu, Haoxiang Zhang, Luoyi Fu
Artificial Intelligence · Cross-listed from cs.CV
UniqueSplat: View-conditioned 3D Gaussian Splatting for Generalizable 3D Reconstruction
Haixu Song, Xiaoke Yang, Shengjun Zhang et al. (+2 more)
In this paper, we propose UniqueSplat, a view-conditioned feed-forward 3D Gaussian Splatting model to reconstruct customized 3D radiance fields for each view query. Existing feed-forward methods such as pixelSplat and MVSplat aim to generate fixed Gaussians across all views of each scene by minimizing the error between rendered views and ground-truth images. However, such fixed Gaussians generally render images from all views and lack the ability to adapt to specific viewpoints, as they do not incorporate target view information when predicting Gaussians. To address this, our UniqueSplat learns the view-conditioned information as a prior and incorporates this knowledge into network parameters, so that Gaussians are dynamically adjusted in accordance with different views. Specifically, we propose a two-branch view-conditioned hyperNetwork to simultaneously learn view-agnostic
embeddings and view-specific knowledge, which not only explores the shareable knowledge from various views, but also adapts the model to specific views at test time. Extensive experiments on widely-used
datasets including RealEstate10K, ACID and DTU demonstrate the superiority of UniqueSplat over the state-of-the-art methods. Moreover, UniqueSplat encouragingly outperforms existing methods in cross-
dataset evaluation, showing its notable generalization ability.
Read the full abstract ▾
All authors Haixu Song, Xiaoke Yang, Shengjun Zhang, Jiwen Lu, Yueqi Duan
Artificial Intelligence
Beyond Solution-Centric Search: Adaptive Inquiry and Knowledge Revision for Autonomous ML Engineering
Shaokang Fu, Yulong Tao, Linbo Jin et al. (+7 more)
Long-horizon autonomous research tasks such as machine learning engineering require systems to make interdependent decisions under a limited budget. Existing LLM-based
agents typically organize candidate-solution improvement through tree, graph, or chain structures, meaning that the search process determines how information is acquired and managed. We call this design solution-centric search and propose instead the information paradigm, in which an evolving information state represents the system's understanding of the task and guides solution improvement. We instantiate this paradigm in Iris, an inquiry-revision loop. For information acquisition, Iris generates local action plans from the current information state and uses epistemic actions to probe decision-critical unknowns without modifying the retained solution. For information management, Iris synthesizes observations across experiments into task knowledge composed of revisable claims with explicit scope and status. It updates this knowledge as new evidence arrives and constructs each decision context from raw evidence, structured summaries, or task knowledge at the required level of detail. On MLE-Bench, Iris attains a 64.9% any-medal rate under a 12-hour budget, the highest among compared systems. Across four tasks spanning harness engineering and model post-training, Iris also demonstrates cross-domain generalization.
Read the full abstract ▾
All authors Shaokang Fu, Yulong Tao, Linbo Jin, Jiarong Zhao, Qiming Shi, Tianjun Pan, Haonan Li, Chengyu Wang, Jia Wu, Chengfu Huo
Artificial Intelligence · Cross-listed from cs.CL
Self-Improving Large Language Models via Progressive Experience Evolution
Shijie Ren, Xiting Wang, Meng Li et al. (+8 more)
Large
language models (LLMs) capable of self-improvement require not only effective policy optimization, but also a principled mechanism for transforming transient interaction experience into persistent model capabilities. Existing self-improvement paradigms remain fragmented: test-time methods can explicitly extract experience but cannot internalize it into model parameters, whereas training-time optimization methods can update model parameters but lack an explicit mechanism for accumulating transferable experience. Bridging these two paradigms requires a critical intermediate stage that remains underexplored, namely experience
distillation. To address this gap, we propose SPEE (Self-Progressive Experience
Evolution), a unified post-training framework that sequentially performs explicit experience
evolution followed by implicit policy optimization. During explicit experience
evolution, SPEE reflects on trajectories collected from multiple interactions to extract, verify, and progressively evolve transferable experience, which is subsequently internalized into the policy through privilege-guided On-Policy Self-
Distillation (OPSD). During implicit policy optimization, reward-driven reinforcement learning leverages these internalized priors to explore novel solution strategies. In the experience
evolution stage, a continuously evolving global experience pool consolidates knowledge from both successful and failed trajectories, filters out low-utility experience, and mitigates post-hoc rationalization induced by individual trajectories. Experiments on five mathematical
reasoning benchmarks demonstrate that SPEE consistently outperforms both test-time and training-time self-
evolution baselines across three model scales. The source code is available at https://github.com/rrrsj/SPEE.
Read the full abstract ▾
All authors Shijie Ren, Xiting Wang, Meng Li, Yujie Guo, Yunhang Yao, Ziheng Peng, Xunlong Wang, Yuetan Chen, Haoyang Zhou, Yunlong Liang, Fandong Meng
Artificial Intelligence
MemArbiter: Decision-Time Memory Arbitration for Long-Horizon LLM Agents
Jiajun Dong, Yutao Hu, Fengrui Fan et al. (+3 more)
Large
language model (LLM)
agents must retain and use cross-step information to act coherently in long-horizon tasks. Existing methods improve
memory accessibility, yet action-relevant information may still fail to guide the current decision because it is poorly formed, organized, prioritized, or presented. We call this post-access failure the
Memory-Action Gap. We propose MemArbiter, a function-aware
memory arbitration framework that addresses the
memory-management-induced component of this gap. MemArbiter decomposes interaction histories into atomic items, organizes them into five functional
Memory Banks, and combines bank-level demand, item-level relevance, focal-ambient representations, and a temporal presentation gate to dynamically control
memory salience. We evaluate MemArbiter on ALFWorld against Flat
Retrieval and Flat Recency under unified per-step
memory budgets. With an open-weight action-generation model, MemArbiter achieves success rates of 82.8% and 92.5% under 500- and 750-
token budgets, outperforming the strongest baseline by 20.9 and 25.4 percentage points, respectively. It also improves post-failure recovery and reduces failed-action repetition and state-action recurrence. These results show that function-aware
memory arbitration enables accessible information to guide actions more effectively.
Read the full abstract ▾
All authors Jiajun Dong, Yutao Hu, Fengrui Fan, Shihan Dou, Yueming Wu, Deqing Zou
Artificial Intelligence · Cross-listed from cs.CL
IACM-RL: Intent-Aware Context Management and Reinforcement Learning for Complex Tool Invocation under Dynamic Intent Fluctuations
Dingwei Zhu, Jiahan Li, Chengjun Pan et al. (+22 more)
Executing long-horizon tool invocations in real-world environments is severely challenged by dynamic user intent noise. Existing methods attempt robustness via implicit history scanning or text compression, yet predominantly assume perfect instructions in simplistic scenarios. Inevitably, under fluctuating contexts, obsolete constraints dilute model
attention, triggering intent deviation and repetitive API loops. To resolve this, we propose IACM-RL, a comprehensive framework for robust tool invocation. First, we introduce the DynamicIntent pipeline, synthesizing trajectories across several fine-grained fluctuation scenarios, paired with a five-dimensional diagnostic metric suite. Second, IACM-RL deploys a BeliefState-based Self-Generated Context Manager that proactively tracks shifting goals and isolates overwritten parameters using structural stale flags. To autonomously internalize this state-tracking capability, we optimize the policy using a hierarchical intent-driven reward alongside two auxiliary losses (CM extraction and state
distillation). Experiments on DynamicIntent, BFCL-V3, and $\mathrm{\tau}^2$-Bench demonstrate that IACM-RL outperforms baselines, reducing infinite loops and stale context errors while enhancing out-of-domain generalization.
Read the full abstract ▾
All authors Dingwei Zhu, Jiahan Li, Chengjun Pan, Yunxian Yang, Yunbin Zhao, Yunke Zhang, Zhonghang Lu, Zhuohui Sheng, Chenhao Huang, Jiahang Lin, Yajie Yang, Junlin Shang, Shichun Liu, Yuhui Wang, Honglin Guo, Junjie Ye, Xin Guo, Jiazheng Zhang, Ming Zhang, Shihan Dou, Zhiheng Xi, Tao Gui, Qi Zhang, Xipeng Qiu, Xuanjing Huang
Artificial Intelligence · Cross-listed from cs.SD
Uncertainty-Aware Crossmodal Fusion for Classification of Animal Behavior
Ehsan Yaghoubi and Florian Haselbeck
Artificial intelligence offers substantial potential for acoustic monitoring of animals, from welfare assessment in precision livestock farming to wildlife conservation and ecological research, where vocalizations can indicate health, stress, and social states earlier and at lower cost than manual observation. However, recordings in these settings are obtained under uncontrolled conditions, including environmental noise, reverberation, overlapping calls, and sensors that degrade without notice. As a consequence, automated classification of animal vocalizations remains challenging, and the two dominant acoustic representations show complementary limitations: raw waveforms preserve temporal microstructure but degrade under clipping and reverberation, while log-Mel spectrograms capture harmonic organization but lose phase information and are sensitive to broadband noise. To address these challenges, we propose Uncertainty-Aware Fusion (UAF), a dual-stream framework that estimates Gaussian uncertainty for each representation and fuses them via uncertainty weighting. This mechanism assigns greater weight to the more confident representation with no reliability labels required. In a cross-species, identity-based evaluation excluding all individuals seen during training, UAF (mean pooling) achieves 59.4% accuracy / 39.7% macro F1 on the 17-class SoundWel pig vocalization
benchmark and 73.1% accuracy / 71.5% macro F1 on the 3-class DogBark
dataset, outperforming static-concatenation fusion by 15.7% and 20.4% relative macro F1, respectively. Ablations over four temporal aggregation strategies show that uncertainty fusion, rather than the temporal characteristics of animal calls, is the primary driver of the performance gain.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.AR
DeGS: A Scalable 3DGS Architecture via Decoupled Workload Parsing and Reorganization
Minnan Pei, Gang Li, Zeyu Zhu et al. (+7 more)
3D Gaussian Splatting (3DGS) has emerged as a leading technique for real-time novel view synthesis, yet existing 3DGS accelerators suffer from poor architectural scalability: increasing the number of PEs leads to marginal performance improvement during rendering. We identify that the root cause is the tightly coupled “checking-while-blending” dataflow, which exacerbates PE underutilization caused by spatial redundancy from irregular Gaussian coverage and temporal redundancy from asynchronous pixel-wise termination under parallel execution. To address this issue, we propose DeGS, a scalable architecture for efficient 3DGS inference. To systematically eliminate the redundancies inherent in rendering, DeGS exploits a decoupled dataflow, restructuring the coupled $\alpha$-checking, transmittance checking, and $\alpha$-blending of the standard rendering process into consecutive workload parsing, reorganization, and blending stages. This allows the fragmented, length-variable, and temporal-dependent workloads to be reorganized into compact, conflict-free, and dense workloads prior to blending, thereby significantly improving PE utilization during parallel blending. Implemented in 28 nm technology, DeGS achieves 2.36$\times$–7.25$\times$ throughput, 1.82$\times$–6.02$\times$ end-to-end speedup, and 1.59$\times$–4.42$\times$ energy efficiency over state-of-the-art 3DGS accelerators (GSCore, GBU, GCC) across diverse scenes and resolutions (720p to 8K). Moreover, scaling from 16 to 1024 PEs, DeGS maintains over 80% PE utilization at high resolutions, significantly outperforming existing accelerators.
Read the full abstract ▾
All authors Minnan Pei, Gang Li, Zeyu Zhu, Siting Wang, Junwen Si, Zhuoran Song, Yu Feng, Fangxin Liu, Xiaoyao Liang, Jian Cheng
Artificial Intelligence
Fetch-then-Explore: Decoupling Selection from Extraction over a Persistent Workspace for Search Agents
Qi Liu, Yiqun Chen, Zidan Chen et al. (+6 more)
Search
agents now answer questions that take dozens of searches to settle, yet how such an
agent reads a page has drawn far less
attention than how it finds one. Nearly all of them use one of two document interfaces, and both tie a page to the moment it is opened. Visit-and-read injects a reading of the page into the message history at fetch time, fixing that reading before the
agent knows which fact it will need. Stateful browsing instead extracts on demand from the page in hand, but holds one page at a time and releases it as soon as the
agent opens another. Either way, a page that turns out to matter many turns later has to be fetched and rendered into context all over again. We propose Fetch-then-Explore, which separates page selection from evidence extraction and keeps what it selects: pages are recorded in a per-question workspace on the filesystem rather than the context window or a transient session, and evidence is pulled from them on demand later. Selection becomes almost free, extraction can wait until the
agent knows what to look for and be repeated as its hypothesis sharpens, and pages are not released when the
agent moves on, so evidence accumulates across the trajectory. In a unified ReAct harness with fixed search, we compare Fetch-then-Explore against snippet-only, visit-and-read, and browsing baselines on two open-web
benchmarks, BrowseComp and WideSearch, across three
agent backbones. It leads BrowseComp accuracy at every backbone and generally matches or exceeds the baselines on WideSearch, and a behavioral analysis traces the gains to the workspace's defining move: returning to a page after leaving it, which it does far more than any transient interface, so evidence missed on a first pass can still be recovered later.
Read the full abstract ▾
All authors Qi Liu, Yiqun Chen, Zidan Chen, Yan Gao, Yi Wu, Yao Hu, Jiaxin Mao, Fengbin Zhu, Tat-Seng Chua
Artificial Intelligence · Cross-listed from cs.LG
How Much Does a Reasoning Summary Reveal? An Observability Ladder for Large Language Models
Andres Algaba, Francesca Carlon, Lynn Delcon et al. (+3 more)
Large
language models often show users a final response and a short
reasoning summary while the full
reasoning trace stays hidden. We introduce an observability ladder that holds each completed run fixed and varies only what a reader inspects to judge whether the answer is correct: the response, a self-summary the model writes from the trace, the trace itself, and internal signals, each with and without the
prompt. Across three
benchmarks and five open-weight Qwen3 and gpt-oss models, we train matched linear correctness predictors on each access level. Without the
prompt, summaries carry most of the trace's ranking signal (mean AUROC 0.774 versus 0.813) and add +0.156 over the response alone. With the
prompt visible, the summary's gain collapses to +0.019, while the trace still adds +0.041. Even at equal length, the trace's last words predict correctness as well as summaries, or slightly better, and carry denser and more discriminative uncertainty and self-correction cues. On MMLU-Pro questions with both correct and incorrect runs, linear summary readers are near chance and trace readers retain only modest signal, both with and without the
prompt (
prompt-withheld AUROC 0.503-0.545 versus 0.544-0.590). With the
prompt withheld, a GPT-5-mini reader recovers substantially more signal from both summaries and traces on gpt-oss-20b, and even then the trace keeps a small +0.034 advantage. Much of the linear readers' trace signal is associated with length. In the common case where users already hold the
prompt, summaries are less helpful than the full trace for monitoring correctness. Monitorability is thus a joint property of the display and the reader, so any monitorability claim, including for faithfulness, should specify both.
Read the full abstract ▾
All authors Andres Algaba, Francesca Carlon, Lynn Delcon, Marthe Ballon, Bert Verbruggen, Vincent Ginis
Artificial Intelligence · Cross-listed from cs.LG
An AI-Based Decision-Support Pipeline for Day-Ahead Photovoltaic Forecasting
Fariba Dehghan, Sebastian Stein, Vahid Yazdanpanah et al. (+2 more)
Reliable photovoltaic (PV) forecasts can support low-carbon energy systems, but deployed sites may have only short and incomplete records. Physical and hybrid methods can be sensitive to weather inputs, calibration, and timestamp-
alignment, while individual machine learning models may capture different parts of the
forecasting problem. We study hourly day-ahead PV
forecasting at a United Kingdom charging station using one year of inverter measurements, with 9.25% of hours missing. The pipeline checks timestamp-
alignment, derives
solar and clearness features, adds short-term weather context, and combines five complementary models using non-negative least squares stacking, with the combination fitted only on validation observations. We compare against smart persistence, a weather-scaled baseline that carries the previous day's PV behaviour forward using target-day irradiance. With retrospective weather, the combined model reduces daylight normalised root mean square error (RMSE) by 31.2% under random day-fold evaluation and by 2.9% under rolling-origin evaluation, although the latter improvement is not robust across days. It also improves by 3.0% over the single model selected from validation performance. Replacing retrospective weather with a public product sampled at a constant 24-hour lead increases daylight RMSE by 13.1% and 4.2% under the two protocols, while retaining positive skill over smart persistence.
Read the full abstract ▾
All authors Fariba Dehghan, Sebastian Stein, Vahid Yazdanpanah, Stephanie Gauthier, Masood Nazari
Artificial Intelligence
Instruction-Conditioned Exploration for Reinforcement Learning with Self-Distillation to an Unconditioned Policy
Jim Dilkes, Vahid Yazdanpanah and Sebastian Stein
Post-training Large
Language Models (LLMs) with Reinforcement Learning (RL) has become an important tool for improving model capabilities, but the LLM action-space structure introduces challenges distinct from classical RL, with implications for inducing exploration. New methods are required that leverage the broad knowledge and flexibility of pre-trained LLMs to deliberately generate diverse experience at training time. We propose Instruction-Conditioned Exploration (ICE), which appends one of a small fixed set of instructions to task
prompts during training, using the same set for every problem, increasing the coverage of behaviours attempted. To facilitate ICE, we combine RL on the instruction-conditioned policy with self-
distillation of its correct rollouts into the unconditioned test-time policy. ICE with this objective improves Qwen3-1.7B held-out pass@1 performance at 4K response length on mathematical
reasoning tasks by $5.0\%$ relative to training with DAPO, with improvement persisting at a longer 8K context. The improvement does not appear for Qwen3-4B at 4K, where the instructions do not expand base-model coverage.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
Geometry-Guided Layerwise FFN Width Allocation in Transformers
Timur Mudarisov, Mikhail Burtsev and Radu State
Feed-forward networks (FFNs) account for a large fraction of
Transformer parameters, yet their hidden width is usually constant across depth. We ask whether this capacity can instead be allocated from a forward-pass measurement of layer behavior. We view each FFN as transporting a cloud of
token representations and quantify the induced
geometric change using correspondence-preserving shift, Gromov-Wasserstein distortion, and degree-one persistent homology under raw and scale-normalized metrics. A layerwise approximation surrogate yields an exact fixed-budget optimizer. Across seven pretrained
language models, raw Euclidean work largely tracks residual-norm growth, whereas normalized work is predominantly front-loaded. Gromov-Wasserstein work is more consistently associated with perturbation-based layer sensitivity than the finite-sample topological estimate. In paired 128M and 256M training runs, several normalized-work schedules reduce mean validation loss relative to both uniform width and a hand-designed cosine taper. With the amplified paired differences at 440M, the best geometry-based allocations improve over uniform substantially larger than the cosine taper, while the anti-topological raw control is worse than uniform.
Read the full abstract ▾
Artificial Intelligence
Cross-Fitted Residual Utility for Primary-Preserving Cognitive Decision Correction in Automatic Modulation Classification
Linzhuo Han, Zongyong Cui and Houbiao Li
Automatic modulation classification research has largely emphasized representation accuracy, but a cognitive receiver must also decide when heterogeneous evidence justifies overriding a trusted default prediction. We study this post-inference problem through cross-fitted residual utility and a primary-preserving cognitive decision policy. A structured KAN-
Fourier classifier supplies the default probability, while neural and non-neural candidates provide observable evidence. Candidate-specific residual utility is learned from train-split out-of-fold predictions, and a disjoint validation split freezes action thresholds, approved transitions, conditional routes, and a unified risk mask before held-out evaluation. On RMLA, RMLB, and HISAR, the complete system improves overall accuracy from 63.632% to 66.332%, 65.161% to 66.168%, and 77.769% to 79.867%, respectively. Controlled comparisons show that the isolated utility target does not uniformly dominate alternative out-of-fold meta-learners; the consistent gain comes from the complete evidence-and-action policy. Paired bootstrap and Holm-corrected McNemar analyses support the controlled gains. A frozen-policy stress test under carrier-frequency offset, I/Q imbalance, and synthetic Rayleigh/Rician fading yields positive gains in all 11 conditions, with every paired 95% confidence interval above zero.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
TBSG-Net: Temporal Bipartite Scene Graph Network for Fine-Grained Video Moment Retrieval
Ji Huang, Yongsheng Dai, Tianyu Ren et al. (+2 more)
Recent advances in proposal-free
Video Moment
Retrieval (VMR) have highlighted the effectiveness of Static Scene Graphs (SSGs). By modeling objects and their relations at the frame level, SSGs enrich
retrieval-oriented
video representations. However, integrating SSGs into VMR remains constrained by two inherent limitations: (1) Lack of Temporal Dynamics. SSGs fail to model how objects and their relationships evolve over time, leading to the loss of essential temporal dependencies in
video representation; and (2) Lack of Explicit Temporal Span Encoding. SSGs do not explicitly encode the duration of relationships, making precise localization challenging. To address these limitations, we propose Temporal Bipartite Scene Graph Network (TBSG-Net)—to the best of our knowledge, the first Dynamic Scene Graph (DSG) based proposal-free VMR model. Specifically, TBSG-Net leverages DSGs to extract event-centric graph representations of the input
video, enabling the modeling of object interactions over time and thus addressing limitation (1). These DSGs are then processed by a novel Dynamic Scene Graph
Embedding (DSG-E) module to capture both Temporal Span and spatio-temporal information. First, DSG-E utilizes a TBSG Constructor to transform DSGs into TBSGs, explicitly encoding objects, relationships, and time spans to tackle limitation (2). Second, the resultant TBSGs are passed into a hybrid TBSG Encoder that integrates a
Transformer variant for global event modeling and a Graph Convolutional Network for detailed relational
reasoning, ultimately producing a more comprehensive spatio-temporal representation. Our experiments demonstrate substantial improvements of TBSG-Net over all baselines.
Read the full abstract ▾
All authors Ji Huang, Yongsheng Dai, Tianyu Ren, Barry Devereux, Hui Wang
Artificial Intelligence · Cross-listed from cs.CL
TextNCA: Neural Cellular Automata for Language Modeling via Hierarchical Local Attention
Avni Mittal, Avinash Anand, Ashutosh Kumar et al. (+7 more)
Can a strictly local, iterated, weight-shared computation primitive support
language modelling, and which of those three properties actually drives the model's behaviour? We define TextNCA, a 1D causal windowed-
attention realisation of the Neural Cellular Automaton primitive, and study a hierarchical variant that cascades three stages with windows $w \in \{8, 32, 128\}$ and $T_s$ shared-weight iterations per stage, all on WikiText-103 at roughly 30M parameters and 60k training steps. The model does not match a parameter-matched
Transformer at this scale (Hier-TextNCA $60.3$ vs.\
Transformer-6L $52.8$ and
Transformer-12L $44.7$ PPL), so we treat it as an analytical probe rather than a proposed alternative. The behaviour we observe is largely explained by the staged narrow-to-wide schedule: a non-iterating sliding-window
Transformer that reuses the same schedule comes within $+4.1$ PPL of the iterated model, while reversing, flattening, or breaking the monotonic ordering of the schedule costs between $+16.7$ and $+70.8$ PPL. Iteration adds a smaller bounded benefit on top of the schedule, with a clear optimum at $T_s{=}4$ and a U-shaped degradation beyond it. The GRU gate and learned per-step
embeddings are required for that benefit to appear, and training with random $T_s$ yields an inference-time iteration-count knob at the cost of substantially higher absolute PPL. We position the work as a controlled reading of which parts of NCA-style computation carry the weight in
language modelling.
Read the full abstract ▾
All authors Avni Mittal, Avinash Anand, Ashutosh Kumar, Dikshant Kukreja, Kritarth Prasad, Sushane Dulloo, Erik Cambria, Timothy Liu, Zhengkui Wang, Rajiv Ratn Shah
Artificial Intelligence · Cross-listed from cs.CL
CompanionBench: A Theory-Anchored, Real-World-Grounded Benchmark for AI Emotional Companionship
Yao Liu, Guangjia Chai, Yuming Huang et al. (+3 more)
LLM companions are deployed at scale in personally consequential settings, yet poorly evaluated. Existing
benchmarks use hand-authored scenarios and prompted simulators, aggregate empathy into one score, and overlook judge biases such as same-family favoritism and scale drift. We introduce CompanionBench, an interactive bilingual
benchmark. To our knowledge, it is the first companion
benchmark to ground both its scenarios and a trained user simulator in de-identified real-world data. A hidden disclosure gate branches each persona's trajectory on the
agent's own behavior, controlling the interaction state space without scripting dialogue. We operationalize ten capabilities derived from 25 theories across psychology and counseling, four of them not graded explicitly by prior work: holding ambiguity, selfobject responsiveness, positive resonance and calibrated challenge.
Agents are assessed on two complementary axes: a subjective ten-capability rubric and a deterministic measure of whether deeper disclosure was earned. A cross-family panel dilutes same-family favoritism; an Item Response Theory model separates
agent quality from judge severity. Theory fixes what to measure and how personas are structured; real data supply events, history, and profiles – coverage from theory, authenticity from data. Rankings are reproducible in both
languages (rho = 0.996 ZH / 0.953 EN). Evaluating 28
agents reveals capability-level differences obscured by aggregate scores.
Emotion regulation and calibrated challenge remain common weaknesses; holding ambiguity discriminates most. Role-play
agents rank near the bottom: immersion does not imply relational competence. Across
agents, the dominant failure mode is substituting surface warmth for substantive relational support. We will release 500 Chinese-English parallel pairs and the evaluation code.
Read the full abstract ▾
All authors Yao Liu, Guangjia Chai, Yuming Huang, Jihao Huang, Lei Wang, Junchen Wan
Artificial Intelligence
HPFA: Hypergraph-Based Paired Failure Attribution for LLM Reasoning
Runchuan Zhu, Hongbin Lai, Bowen Jiang et al. (+4 more)
Reflection is a powerful mechanism for LLM
reasoning, yet its effectiveness hinges on accurately attributing failures to specific
reasoning steps, a capability that current models notably lack. Existing failure attribution methods either require expensive step-by-step counterfactual testing that scales poorly with trajectory length, or treat
reasoning traces as flat sequences that ignore the inherent non-linear logical dependencies. We propose a hypergraph-based paired failure attribution (HPFA) framework that attributes the failure root cause by comparing the hyperedges of the targeted failure
reasoning path against a reference successful path. By reducing the search space, our method efficiently localizes root causes and enables scalable synthesis of attribution data for training a lightweight attributor model via supervised fine-tuning and reinforcement learning. Experiments on mathematical
reasoning and agentic coding tasks demonstrate that HPFA can dramatically increase attribution accuracy and efficiency, and the trained attributor consistently improves
reasoning accuracy at test time, outperforming baselines that lack graph structure or paired analysis.
Read the full abstract ▾
All authors Runchuan Zhu, Hongbin Lai, Bowen Jiang, Junrui Zhang, Zhangheng LI, Ostap Kilbasovych, Junyuan Hong
Artificial Intelligence
EduZone: A Framework for Evaluating LLM Safety for K-12 Students and Teachers
Junyeong Park, Jieun Han, Haneul Yoo et al. (+3 more)
Large
language models (LLMs) are increasingly used across diverse tasks in K-12 education, yet existing safety evaluations rarely examine how harmful or inappropriate content appears in interactions between LLMs and students or teachers. To address this, we present EduZone, an evaluation framework for LLM safety across diverse educational scenarios. Our framework systematically combines (1) student- and teacher-facing LLM usage contexts, (2) fine-grained curriculum concepts, and (3) 6 risk categories and 28 subcategories spanning both conventional and education-specific harms to generate contextually grounded
adversarial interactions. We construct these interactions in three settings: single-turn requests, static multi-turn conversations, and dynamic multi-turn conversations. Using these interactions, we evaluate ten LLMs using four safety levels: refusal, safe assistance, risky assistance with safety guidance, and fully risky assistance. Our results reveal greater vulnerability to education-specific risks and dynamic multi-turn interactions, while existing safety guardrails fail to adequately address these risks. EduZone advances LLM safety in education by providing an automated, scalable evaluation framework that supports the development and deployment of safer LLMs in K-12 education.
Read the full abstract ▾
All authors Junyeong Park, Jieun Han, Haneul Yoo, So-Yeon Ahn, Jinsung Yoon, Alice Oh
Artificial Intelligence · Cross-listed from cs.RO
MANGO-Grasp: Mahalanobis Fields over Geometry-Oriented 3D Gaussians for Cross-Embodiment Dexterous Grasping
Heng Zhang, Kevin Yuchen Ma, Mike Zheng Shou et al. (+2 more)
Cross-embodiment dexterous grasping aims to synthesize stable grasps across heterogeneous multi-fingered hands with little or no embodiment-specific tuning. Existing interaction-centric methods achieve promising results, but their object representations often underrepresent local surface geometry, hand descriptors do not explicitly capture both morphology and kinematics, and hand-object interaction modeling is typically isotropic. We propose MANGO-Grasp, an anisotropic interaction framework that represents objects as geometry-oriented 3D Gaussian primitives to better preserve local surface geometry, and encodes hands into pretrained descriptors that capture both morphology and kinematics. Hand-object interaction is modeled using anisotropic Mahalanobis fields that reflect the directional nature of contact and guide a shared optimization for grasp realization across embodiments. On the CMAP and MultiGripperGrasp
benchmarks, MANGO-Grasp outperforms the strongest seen-hand baseline by up to 8.24 percentage points in simulation. It also transfers zero-shot to the unseen SharpaWave hand, improving over the strongest zero-shot baseline by up to 16.57 percentage points, and achieves 86% success in real-world experiments. The code and additional materials will be made available upon publication at https://connor-zh.github.io/MANGO-Grasp/.
Read the full abstract ▾
All authors Heng Zhang, Kevin Yuchen Ma, Mike Zheng Shou, Weisi Lin, Yan Wu
Artificial Intelligence
Before Reasoning Can Fail: Pre-Evidence Procedural Failures in Agentic RAG
Daeyoung Roh and Donghee Han
Agentic
retrieval-augmented generation (RAG) systems can fail before evidence-conditioned
reasoning is tested: an
agent may retrieve candidate snippets but finalize without inspecting them. We study this failure mode as a procedural property of the
agent trajectory, decomposing wrong answers into pre-evidence discipline failures and post-gold-read failures using saved tool-call traces, retrieved evidence, read passages, and final answers. Across 12,000 paired trajectories on HotpotQA, 2WikiMultiHopQA, and MuSiQue, the two failure types are largely non-redundant: the both-trigger rate is in [11.2%, 13.1%] across regex and spaCy entity extractors. We then evaluate Read-Gate, a minimal runtime invariant requiring an
agent to read after search and before finalization. Forced reading improves LLM-Acc by 14.9-19.9 points on trajectories that would otherwise skip reading and by 3.2-9.4 points on full minimal-
reasoning cells. Additional diagnostics show that larger hidden thinking budgets do not necessarily increase evidence inspection. Together, these results indicate that evidence-gathering should be evaluated as a trajectory-level control problem, separately from answer-side
reasoning.
Read the full abstract ▾
Artificial Intelligence
HALT: Verification-Aware Stopping for Retrieval-Augmented Search Agents
Daeyoung Roh and Donghee Han
Retrieval-augmented search
agents answer multi-hop questions by repeatedly issuing search queries and accumulating evidence. This creates a stopping problem: after the necessary evidence has appeared, further
retrieval often adds cost, latency, and distracting context rather than useful information. We frame stopping as evidence coverage rather than generator confidence, and introduce HALT, a lightweight verification-aware policy that leaves the search
agent unchanged. Given expected hop claims, HALT stops only when cumulative evidence supports each required claim. Across three multi-hop QA
benchmarks, HALT reduces redundant search while largely preserving exact match. We separate a deployable setting, where hop claims are generated from the question, from a diagnostic upper bound that uses gold supporting-fact annotations: generated claims give smaller but still exact-match-preserving savings, while gold claims show the larger savings available when hop targets are clean. Baseline comparisons and ablations show that this behavior is driven by claim-evidence
alignment rather than generic sufficiency, fixed stop positions, or lexical overlap. Open-corpus pilots further suggest that HALT abstains when coverage cannot be reliably verified. Overall, evidence coverage provides a practical runtime control signal for improving
retrieval-augmented
agents without retraining or modifying the host
agent.
Read the full abstract ▾
Artificial Intelligence
Evolving in the Agent Jungle via History-Informed Opponent Awareness
Zhaofeng Zhang, Linhan Xia, Rui Liu et al. (+3 more)
Learning to adapt strategies through interaction is a key step toward more general and autonomous LLM
agents. Existing approaches typically achieve behavioral adaptation by revising skill libraries. However, in multi-
agent environments, opponents may simultaneously update their strategies, causing the environment itself to evolve continuously. Applying skill-revision methods designed for static environments in such settings therefore amounts to updating against an obsolete reference. To address this challenge, we introduce OASE (Opponent-Aware Selective
Evolution), which identifies and adopts genuinely beneficial skill revisions in dynamic multi-
agent environments. Specifically, OASE conducts paired comparisons between a candidate skill and the incumbent under identical conditions anchored by historical snapshots of opponent strategies, and adopts the candidate only when its estimated payoff gain exceeds an acceptance threshold. We evaluate OASE in two decision-making scenarios: first-price auctions and private-cost Cournot competition. Experimental results show that, compared with a Reflexion-style baseline, OASE achieves a lower final
equilibrium distance in both environments while accepting substantially fewer skill revisions, thereby suppressing strategy changes that lack sufficient payoff support. OASE therefore replaces blind updating with evidence-anchored selection, allowing
agents to adapt stably and efficiently even as opponents continuously evolve.
Read the full abstract ▾
All authors Zhaofeng Zhang, Linhan Xia, Rui Liu, Yihao Wang, Binrui Shen, Shengxin Zhu
Artificial Intelligence · Cross-listed from cs.DC
TALSC: Timeliness-Aware Large-Small VLM Collaboration for Infrastructure-Assisted Autonomous Driving
Mengmeng Zhu, Yuxuan Sun, Wei Chen et al. (+1 more)
The deployment of
Vision-
Language Models (VLMs) in autonomous driving (AD) systems is constrained by on-board computing power, restricting vehicles to small VLMs (SVLMs) with limited perception and
reasoning capabilities. Infrastructure-assisted AD alleviates this resource constraint by enabling collaboration with large VLMs (LVLMs) at edge servers. However, in dynamic vehicular environments, the utility of sensory data for downstream tasks decays rapidly, making timeliness of information a critical concern. To balance the accuracy gains of LVLMs with their latency-induced timeliness degradation, we develop a Timeliness-Aware Large-Small VLM Collaboration (TALSC) framework. Specifically, we first model the Age of Information (AoI)
evolution for VLM inference and characterize the coupling among AoI,
token length, and task performance to formulate a general timeliness metric. Building on this, we propose the TALSC online scheduling
algorithm. Since scheduling decisions have a delayed impact on future timeliness metric and the output
token number is unknown at scheduling time, we design a Lyapunov drift-plus-estimated-penalty
algorithm and provides a guaranteed performance. In simulation, we first conduct a case study to derive a fitted timeliness metric based on nuScenes
dataset, and further show that TALSC outperforms baselines under various communication and computing settings, achieving up to a 12.6% normalized improvement in Micro-F1 score compared with the best-performing baseline.
Read the full abstract ▾
All authors Mengmeng Zhu, Yuxuan Sun, Wei Chen, Bo Ai
Artificial Intelligence
Long-Horizon Autonomous Architecture Research with a Language-Model Agent: A Behavioural Case Study
Aon Safdar and Mohamed Saadeldin
We study what happens when a single general-purpose large
language model acts as the sole researcher on a long-horizon neural architecture design problem. The
agent receives a scientific question, an initial hypothesis and motivation, a compute budget, and research affordances (source and experiment management, experiment tracking, literature access, and persistent
memory), then autonomously proposes, implements, evaluates, and records experiments over an extended period. The study comprises three phases, separated by human-declared transitions, that progressively expand the
agent's tool surface or problem scale. Across approximately 100 sequential experiments, the
agent improves a non-standard
Vision Transformer from a weak baseline to a stronger, efficient model on small
benchmarks and a usable but sub-SOTA model on ImageNet-1K, while producing a dense behavioural trace. We report four findings.(i)Productivity exhibits a clear phase structure: rapid early gains, a multi-dozen-hypothesis saturation wall, and recovery, with recovery triggered by expanding the action surface rather than changing the underlying model.(ii)A single early hypothesis contributes more to accuracy gain, with later improvements long-tailed.(iii)The preference for greedy, incremental hypotheses is largely workflow-induced: a commit-or-discard evaluation rule is isomorphic to greedy hill-climbing; the remainder reflects risk aversion after bold failures and anchoring on familiar literature. (iv)The
agent independently rediscovers established results and, in the unfamiliar regime of pure channel
attention, overturns a standard design choice. We conclude that workflow design was at least as influential as
agent capability in this study and propose diversified search, budgeted moonshot hypotheses, explicit forks, and regime-aware re-validation as testable directions for future autonomous research.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
SPARE: Structural Parameter-Free Affinity Regularization for Flow Matching
Zong-Wei Hong, Jinglun Li, Shen Zhang et al. (+3 more)
Denoising
diffusion transformers achieve strong generation quality but converge slowly during training. Regularizing their internal representations has emerged as an effective accelerator, yet existing methods split into two families with complementary costs. Target-based methods strengthen representations by aligning them to external features, which requires an external encoder and a learnable projection head to bridge feature spaces. Target-free methods hold no reference at all, and can only repel the model's own features across samples or layers, discarding whatever structure the data contains. Prior work suggests that spatial structure, rather than global semantics, drives the gains of
alignment. We therefore ask whether such structure can serve as a target directly, and whether it exists not only within an image but across images. Our key insight is that the clean data latent already carries this structure in the relations among its
tokens, where a relation is the similarity between two
tokens, a single scalar comparable across feature spaces without a projection head. We propose Structural Parameter-free Affinity Regularization (SPARE), a regularizer that matches the pairwise affinities of intermediate
tokens to those of the clean latents. To exploit this structure fully, SPARE extends the matching to
token pairs across images, precisely the pairs that prior target-free methods repel by default, and calibrates both relation types with a single learning objective. On ImageNet $256 \times 256$ with SiT backbones under matched 400K-iteration budgets, SPARE adds no encoder, head, or parameters and only 0.08 GB of training
memory, yet attains the lowest FID among parameter-free regularizers in every tested setting, recovers 37 to 54% of REPA's FID reduction, and improves over REPA when combined with it, reaching FID 1.90 under classifier-free guidance at 1M iterations.
Read the full abstract ▾
All authors Zong-Wei Hong, Jinglun Li, Shen Zhang, Yuhan Liu, Linze Li, Yao Tang