跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+817新增论文
+672其中 AI 相关
+98已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-19 06:59 · 全站最近入库:2026-11-19 06:59各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 12:48 — 2026-11-19 12:48(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-19 06:59最近发现新增版本:2026-11-19 06:59库内论文最新更新时间:2026-11-18 08:00

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

26,871 已收录论文22,375 AI 相关8,156 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 28 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-19 06:57(北京时间)
新论文 · v1arXiv:2611.20953

Same Fix, Different Weights: Replicating Weight-Space Programming with Claude

Subham Kundu

原文摘要

Takehi et al. (2026) report that GPT-6 Astra, given the weights of small neural networks as text, can say what they were trained to do, repair them, and write new ones without training. I re-ran the core of that study with Claude Opus 5.5 in ten pre-registered experiments, with untrained-network and permutation controls and hand audits of every sandboxed code cell. Reading was mixed: Claude named 3 of 4 Atari games from policy weights and found the measured copying heads in 15/15 character-LM runs, but its answers about training data often followed priors. Unlike the original, it identified a sycophantic Gemma 3 270M finetune from weights alone in 6/6 runs. One-shot weight edits took a failing MountainCar policy from 0% to 100% success and a hovering LunarLander policy from 0% to 85.6%; zero-shot networks beat PPO at 100k steps on 4 of 5 environments, each implementing a textbook controller. With no training or forward passes, Claude wrote a 1.48M-parameter 9x9 Go network that won 85-94% of games against three PPO opponents. Given Qwen3.5-0.8B's weights and 10, 25 or 50 forward passes of 64 practice problems, Claude raised MATH-500 from 6.7% to 51.4%, 55.2% and 58.1% (+51.4 points at 50 passes, 95% CI [47.9, 55.0]; original 48.1%), GSM8K from 11.6% to 68.2% and IFEval from 22.2% to 54.2%. Unlike the original, which removed an internal loop direction, every edit acted at the output: it scaled the final norm, raising the effective sampling temperature, and lowered the logit of "Wait" while raising that of </think>. The norm scaling alone reaches only 20.9-22.1% on MATH-500. On Gemma 3 270M, which rarely loops, no edit helped (-0.8 points), as in the original.

cs.LG

新论文 · v1arXiv:2611.20401

Interlocking Tiling of 3D Finite Domains

Xuyang Ma, Rulin Chen, Chaewoon Kim, Ying He, Peng Song

原文摘要

A K-hedral tiling of a 3D finite domain is a covering of the domain with tiles without gaps or overlaps, where each tile is congruent to one of the K distinct shapes called prototiles. To use 3D K-hedral tilings in practice, they need to satisfy two requirements. First, K should be small to reduce the fabrication cost (e.g., through reusable molds) and simplify the (dis)assembly process. Second, a K-hedral tiling result should be structurally stable, in particular via interlocking, where all the tiles are immobilized by their geometric arrangement, except the key. Since these two requirements are often in conflict, existing approaches always trade one for the other. In this work, we formulate a new tiling problem called interlocking tiling, aiming to generate 3D K-hedral tilings that satisfy the above two requirements simultaneously. Given a 3D finite domain represented as a voxel grid, our goal is to compute a small prototile set whose instances exactly tile the input domain while ensuring that the tiling result forms an interlocking assembly. To solve this challenging problem, we propose a twolevel computational approach that hierarchically decomposes the input domain into interlocking parts (domain-level partitioning) and further partitions each part into interlocking tiles (part-level partitioning). To guide the part-level partitioning, we propose a new graph representation with regular structures to facilitate the generation of tiles with identical shapes while ensuring that they are interlocking. Our approach allows generating interlocking tilings that exactly cover 3D finite domains with diverse shapes while achieving high prototile reuse. We compared our approach with state-of-the-art methods to demonstrate its superiority, and validated the practicality of our interlocking tiling results via fabricating two of them.

cs.GR

新论文 · v1arXiv:2611.20326

Simulating access to alcohol screening, brief intervention, referral to and receipt of specialty treatment for alcohol use disorder in California and Texas: A study protocol for an agent-based modeling study (AUDTxSim)

Rainer Hilscher, Andy Kawabata, Nick Kruskamp, William Dowd, Nina Mulia, Carolina Barbosa, Georgiy Bobashev, Jessica Duncan Cance 等 11 位作者

原文摘要

Alcohol use disorder (AUD) affects millions of US adults. Primary care-based screening, brief intervention, and referral to treatment (SBIRT) is a leading strategy for identifying heavy drinking early, but access to primary care, screening, and specialty treatment differs markedly across subgroups. Whether scaling SBIRT and expanding treatment capacity narrows or widens these differences is unknown and cannot feasibly be tested in a randomized trial. We present AUDTxSim (AUD Treatment Simulation), a stochastic, spatially explicit, individual-based model of alcohol use, AUD, SBIRT, and specialty treatment for the adult populations of California and Texas. Agents represent every adult in synthetic populations of the two states (28.0 and 19.3 million adults), each geo-located to a Census tract and linked to nearby treatment facilities. In daily time steps, the model simulates: (1) annual transitions among eight alcohol use and AUD states; (2) healthcare visits, screening, brief intervention, and referral; (3) treatment seeking, standalone medication for AUD, mutual help, and requests for hospitalization, residential, or outpatient care; (4) a treatment manager that allocates individuals to facilities by preference, geography, and capacity, with telehealth substitution and waitlists; and (5) treatment progression, stepdown between levels of care, and post-treatment mutual help. It is implemented in Python with Repast4Py to run in parallel on a computing cluster. Once calibrated to state, rural-urban, and demographic targets for healthcare visits, heavy drinking and AUD prevalence, and mortality, AUDTxSim will project population-level and subgroup-specific effects of alternative SBIRT reach and treatment capacity scenarios on AUD prevalence, treatment receipt, and mortality. We report the model structure, decision logic, data sources, and calibration plan.

cs.CY

新论文 · v1arXiv:2611.18307

AI for Parameter Selection in Sample-based Quantum Diagonalization

Doriela Grabocka, Dimitrios Christofidellis, Federico Zipoli, Anna van Elst, Patrick Ruch, Antonio Foncubierta-Rodríguez, Siddhartha Mishra, Matteo Manica

原文摘要

Sample-based Quantum Diagonalization (SQD) methods provide a way to compute molecular ground-state energies, but their accuracy and runtime depend sensitively on configuration choices such as the subspace dimension, propagation timestep, and the allocation of a fixed sampling budget across circuits, for which no formal selection method exists. In this work, we explore whether oracle models can predict the quantum runtime and approximation error from the Hamiltonian and the candidate configuration. Using synthetic Hamiltonians with controllable spectral properties, we show that machine-learned models can map a Hamiltonian directly to the best-ranked configuration within a predefined grid for the SKQD algorithm, validated against classical Krylov propagation, achieving strong rank correlations between true and predicted error and runtime, Spearman $ρ= 0.969$ and $ρ= 0.905$, respectively. Extending the approach to 50 molecular Hamiltonians (20-50 qubits), with reference ground-state energies from classical calculations, we train regressors that predict the simulation error and runtime for a given molecule--configuration pair ahead of execution, and classifiers to distinguish, from 600 SqDRIFT quantum-hardware runs, whether a configuration achieves a ground-state energy within chemical accuracy ($1.6\times10^{-3}$ Ha $\approx 1$ kcal/mol). For the molecular Hamiltonians, the models additionally provide an a priori feasibility and resource estimate, predicting whether a molecule can reach chemical accuracy on current hardware and at what runtime, thus correctly classifying held-out molecules at the molecule level, achieving balanced accuracy 61.8% in-distribution and 100% for five larger held-out molecules.

physics.chem-ph · cs.AI

版本更新 · v2arXiv:2611.14281

UniFlow-RS: Unified Generative Prediction for Heterogeneous Remote-Sensing Tasks

Jianlong He, Liang Liao, Jing Xiao, Chaofeng Chen, Gui-Song Xia

原文摘要

Remote-sensing foundation models increasingly unify representations across sensing modalities and downstream tasks, yet their prediction mechanisms remain largely task-specific. Separate decoders, prediction heads, and objectives leave prediction fragmented across tasks. This naturally raises a question: can heterogeneous remote-sensing tasks share the prediction process? UniFlow-RS answers this question with a single latent flow-matching predictor shared across tasks. This requires addressing two key challenges: unifying heterogeneous prediction spaces and harmonizing joint optimization across tasks. First, we cast heterogeneous task outputs into structured visual targets in a common latent space, allowing them to be modeled by the same pretrained flow predictor. Task-dependent conditioning further accommodates differences in input configurations. Second, we optimize all tasks with a common Flow Matching objective, eliminating task-specific prediction losses. To mitigate optimization mismatch across tasks and modalities, we introduce Task-Modality Adaptive Timestep Sampling (TMAS), which dynamically allocates training timesteps according to the online learning state of each task-modality group. Using a single shared model, UniFlow-RS handles six tasks and thirteen datasets, surpassing recent task-specific state-of-the-art methods on several benchmarks while remaining competitive on the others. Beyond these benchmarks, the same checkpoint demonstrates strong zero-shot transfer across unseen datasets, task settings, occlusion conditions, and near-infrared false-color imagery, without target-specific adaptation. Codes are available at https://github.com/MindVision-AI/UniFlow-RS.

cs.CV

版本更新 · v2arXiv:2605.02615

TRACED-X enables in vivo imaging of cellular microstructure, intrinsic extracellular diffusivity, tortuosity, and membrane permeability in human glioma patients

Joshua K. Marchant, Hong-Hsi Lee, Elizabeth R. Gerstner, Susie Y. Huang, Bruce R. Rosen

原文摘要

The lack of analytical models describing diffusion time dependence at intermediate time scales in complex tissue microstructure limits accurate quantification of extracellular diffusivity and tissue microstructure. Cell membrane water permeability modulates the signal in complex ways and is often disregarded in tumor microstructure models. We introduce TRACED-X, a Monte Carlo-based model that incorporates diffusion time dependence and membrane permeability in cell distributions to quantify pathologically-relevant properties in solid tumors. Monte Carlo simulations of water diffusion in semipermeable spheres were used to train neural network regressors for parameter estimation. A separate neural network was trained to predict the extracellular effective diffusivity, enabling the estimation of extracellular space tortuosity. Diffusion data from eight mixed-grade glioma patients were fitted with TRACED-X and analyzed in conjunction with clinical images. Simulation results demonstrate improved parameter estimation compared to comparator models. TRACED-X enabled the simultaneous in vivo quantification of intracellular volume fraction, cell size distribution, intrinsic extracellular diffusivity, tortuosity, and membrane permeability in glioma patients. Parameter maps show coherence with radiological findings and expected tumor phenotypes. TRACED-X permits the rapid estimate of unique biophysical properties in solid tumors, such as intrinsic extracellular diffusivity, tortuosity, and membrane permeability, in addition to cell microstructure. TRACED-X maps show promise in identifying heterogenous tumor subregions before and after treatment. Neural network implementations of diffusion time-dependence and tortuosity showed behavior consistent with coarse-graining and effective medium theory, respectively. Future work will explore the clinical utility of TRACED-X parameters in additional patients.

physics.med-ph · cs.LG · eess.IV

新论文 · v1arXiv:2611.12634

How Far Does Second-Order Linear Response Predict Quantization Mismatch? Pre-Registered Measurements on Five Language Models

Dezhen Lu, Zixiang Ni

原文摘要

Serving systems choose a weight format for each model and deployment. A low-cost screen predicts the resulting KL divergence from the unquantized model's Fisher information using $\tfrac12Δ^\top FΔ$, the expansion underlying Hessian-aware and second-order post-training quantizers. We assess the reliability of this screen in GPU batches with channel-wise int8, per-tensor fp8 and group-wise int4 on Qwen3-8B, Qwen3-4B, OLMo-2-7B, Qwen3-32B and Qwen3-30B-A3B. Relative predictions hold across a 16-fold range of perturbation scales, int4 group sizes and rounding modes, while absolute predictions are heavy-tailed. A trimmed average brings 10 of 12 cells on Qwen3-4B and OLMo-2 within 15% of measured KL. The MoE model exhibits nonquadratic response at small perturbations, with int8 KL scaling as $α^{0.63}$. Computing per-sequence gradients in bf16 can cause fourfold prediction errors. We trace these errors to arithmetic in the lower half of the backward pass and eliminate them by using fp32 there. Batch-invariant kernels run at 0.55 times the throughput, while changes in reduction order introduce a floor of 0.17-0.22 nats per sequence. We build a screener with one streamed fp32 Fisher pass, trimmed tensor-block predictions with a band, a flag for aligned tensors and per-layer format allocation. On the unseen Mistral-7B-v0.3 family, all six predictions fall within the band, all three KL budgets are met, and vLLM throughput is predicted within 0.4%.

cs.CL · cs.CR · cs.LG

版本更新 · v2arXiv:2505.20536

Covariate-Adjusted Deep Causal Learning for Heterogeneous Panel Data Models

Guanhao Zhou, Yuefeng Han, Xiufan Yu

原文摘要

This paper studies the task of estimating heterogeneous treatment effects in causal panel data models with covariate effects. We propose a novel Covariate-Adjusted Deep Causal Learning (CoDEAL) framework that cohesively deals with the underlying heterogeneity and nonlinearity of both panel units and covariate effects. CoDEAL integrates nonlinear covariate effect components (parameterized by a feed-forward neural network) with nonlinear factor structures (modeled by a multi-output autoencoder) to form a heterogeneous causal panel model. The nonlinear covariate component flexibly captures complex covariate influences on outcomes, and the nonlinear factor decomposition enables CoDEAL to effectively capture both cross-sectional and temporal dependencies inherent in the data panel. This latent structural information is subsequently integrated into a customized matrix completion algorithm, thereby facilitating more accurate counterfactual imputation. Moreover, the use of a multi-output autoencoder explicitly accounts for heterogeneity across units and enhances the model interpretability of the latent factors. We establish theoretical guarantees on the convergence of the estimated counterfactuals and demonstrate the compelling performance of the proposed method using extensive simulation studies and real-world data applications.

stat.ML · cs.LG · econ.EM · stat.ME

版本更新 · v7arXiv:2601.03079

Learning to Diagnose and Correct Moral Errors in Large Language Models

Bocheng Chen, Xi Chen, Han Zi, Haitao Mao, Zimo Qi, Xitong Zhang, Yuda Chen, Kristen Johnson 等 9 位作者

原文摘要

Enabling large language models (LLMs) to diagnose and correct moral errors is crucial not only for ensuring their reliable deployment but also for charting a pathway toward continuous improvement. Existing research has primarily followed two directions: fine-tuning LLMs on corpora of morally appropriate texts and developing carefully designed prompting strategies to promote moral behavior, including moral self-correction and self-improvement. However, both lines of work often remain brittle, relying heavily on shallow heuristics rather than enabling LLMs to reason about why and how a given moral discourse is morally (in)appropriate. In this paper, we address this challenge by developing pragmatic inference-driven approach that help LLMs learn to diagnose and correct moral errors. Pragmatic inference is the reasoning process through which implied meanings are derived, a central concept in linguistics. To improve the reasoning efficiency, our approach adapt the inference procedure according to the inferential load required by different moral discourses. Empirical results demonstrate that our method not only improves LLMs' ability to diagnose and correct moral errors, but also generalizes effectively across tasks without relying on shallow heuristics.

cs.CL

新论文 · v1arXiv:2611.06113

Semantic Prior-Guided Incomplete Cross-Modal Retrieval with Noisy Labels

Jiali Chen, Ruitao Pu, Yingke Chen, Dezhong Peng

原文摘要

Cross-modal retrieval (CMR) aims to bridge the semantic gaps between modalities by learning shared representations. Unfortunately, due to the dynamism and inconsistencies in real-world data acquisition, multimodal datasets frequently encounter severe modality incompleteness. Existing incomplete CMR methods adhere to a label-driven paradigm, relying heavily on categorical annotations as the semantic benchmark. However, label noise is difficult to avoid and often coexists with modality incompleteness. To address this challenge, we propose a novel semantic prior-guided framework for incomplete cross-modal retrieval with noisy labels, termed SPIRN. Specifically, SPIRN uses fixed semantic priors as a consistent reference for robust representation learning, noisy-label handling, and missing-modality completion. Based on this reference, Semantic Priors Guidance (SPG) establishes a stable category-level semantic space, Quality-aware Robust Learning (QRL) dynamically evaluates sample quality to distinguish clean samples from noisy ones and handles noise via reverse learning, Prior-guided Global Label Calibration (PGLC) rectifies corrupted annotations for refined supervision, and Confidence-driven Selective Imputation (CSI) recovers missing modalities from a cross-modal memory bank while filtering unreliable imputations with an increasing threshold. To the best of our knowledge, this is the first framework to jointly tackle the dual challenges of modality incompleteness and label noise in the CMR community. Extensive experiments on four widely used benchmark datasets demonstrate that SPIRN outperforms 11 state-of-the-art methods in various scenarios involving both missing modalities and noisy labels.

cs.CV

版本更新 · v2arXiv:2608.02946

Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model

Yuliang Chen, Weiwei Shi, Rong Zablocki, Animesh Kumar, Jordan A. Carlson, Sheri J. Hartman, Mikael Anne Greenwood-Hickman, Paul R. Hibbing 等 13 位作者

原文摘要

Accurate detection of sedentary behavior is important for studying health risks related to prolonged sitting, but posture-based classification remains challenging with wearable sensors, especially at the wrist. We study whether a deep learning model trained on hip-worn accelerometer data can transfer to wrist-worn accelerometer data for sitting versus non-sitting classification. We use CHAP, a CNN-BiLSTM model originally developed for hip accelerometers, and evaluate its zero-shot performance on wrist data as well as its adaptation through finetuning with varying amounts of labeled wrist data. Experiments are conducted on the iWatch dataset with ground-truth posture labels derived from wearable cameras. The hip-trained model performs strongly on hip data without retraining, but accuracy drops on wrist data due to sensor placement shift. Finetuning CHAP provides consistent advantages over transformer models trained from scratch. These findings suggest that hip-based pretraining provides a useful starting point for wrist deployment, while highlighting the need for wrist-specific adaptation to handle higher signal variability.

cs.LG

新论文 · v1arXiv:2611.05073

Toward Constant-Factor Maximin Share Guarantees for Subadditive Valuations

Bo Li, Ankang Sun

原文摘要

For $n$ agents and $m$ indivisible goods, an agent's maximin share (MMS) is the largest minimum bundle value she can obtain by partitioning the goods into $n$ bundles. Constant MMS approximations are known for additive, submodular, or XOS valuations, but whether such a guarantee exists for general subadditive valuations remains open. We advance this question in two directions. For general subadditive valuations, we exponentially improve the guarantee from $Ω(1/\log\log n)$ to $Ω(1/\log^* n)$, where $\log^* n$ is the iterated logarithm. We then seek constant MMS guarantees under additional structure on subadditive valuations by studying $q$-partitioning valuations: every partition of a bundle into at most $q$ groups admits nonnegative group prices that sum to the bundle's value and whose sum on any collection of groups is at most the value of their union. This hierarchy ranges from subadditivity at $q=2$ to XOS for $q\ge m$. We give the first study of MMS in the intermediate classes and prove constant guarantees whenever $q/n$ has a fixed positive lower bound, including $1/6$ for $q\ge n$ and $1/4$ for $q\ge2n$. For XOS valuations, we further improve the universal guarantee from $4/17\approx 0.2353$ to $5/16=0.3125$. For subadditive valuations, we prove a bidding amplification theorem that gives an exponential bound on each agent's probability of receiving less than half her MMS, while satisfying her budget in every outcome. For $q$-partitioning valuations, we give a scalar criterion that converts bounds on the value lost from disjoint deletions within a bundle into bounds on potential losses, even when an agent contributes zero below the target bundle value. For XOS valuations, we prove a reservation theorem that selects goods for integral assignment together with auxiliary additive valuations and accounts for the loss in rounding.

cs.GT

版本更新 · v3arXiv:2610.38249

Keepace.ai: A Technical Report on a Large Language Model for Sports and Health

Keepace. ai Team, Bowen Wu, Heng Wang, Jufeng Mi, Kang Liu, Kun Li, Long He, Manling Han 等 15 位作者

原文摘要

Large language models hold promise for delivering safe, scientifically grounded, and personalized exercise guidance at scale, yet the sports-and-health domain still lacks curated training corpora and rigorous evaluation benchmarks. Keepace (Keepace.ai) is a domain-specialized LLM built on Qwen3.5-27B through post-training, currently optimized for Chinese-language exercise guidance across three tasks: workout plan generation, sports-and-health knowledge QA, and personal sports data analysis. In internal rubric-based evaluations designed with domain experts, Keepace outperforms its base model and larger general-purpose baselines, with the most pronounced improvements in exercise safety, boundary compliance, structured output completeness, and scientific accuracy. This report summarizes the current design, data pipeline, safety-first evaluation framework, and evaluation results of Keepace.

cs.CL

新论文 · v1arXiv:2611.01864

Quantizing Recurrent Language Models Is Easier Than You Think

Philipp Nazari, Neehal Tumma, Daniela Rus, T. Konstantin Rusch

原文摘要

Linear recurrences are becoming an increasingly common building block in large language models, replacing the growing KV cache of softmax attention with a fixed-size recurrent state that enables efficient decoding. However, the quantization stack has largely been developed around softmax Transformers, whose numerical properties and inference state differ from those of recurrent models. Our goal is to understand these differences and study the design space of quantization methods for recurrent models. In our analysis, we differentiate between two targets. First, extra-state quantization concerns weights and activations outside the recurrent state, whose distributions are shaped indirectly by the sequence mixer during training. In controlled 1.3B parameter models, we find that decayed recurrences develop weaker residual outlier channels than softmax attention and are therefore easier to quantize without incoherence processing. Second, intra-state quantization concerns the recurrent state itself during decode, where repeated updates create a different numerical problem from KV cache quantization. We instantiate our findings as LoReS, a state quantization pipeline combining value-axis scale grouping to match output channel magnitude, decay-based mixed-precision to protect slowly decaying heads or channels, and stochastic rounding to limit error accumulation under repeated requantization. Across Qwen-3.5 9B, Nemotron-H 8B, and Kimi-Linear 48B-A3B, LoReS supports robust 8-bit state quantization and lower effective precision on shorter decode horizons. We implement LoReS in vLLM and demonstrate end-to-end decode speedups.

cs.LG · cs.AI · cs.CL

版本更新 · v5arXiv:2610.27746

S$^2$-MAS: A Society-Scale Multi-Agent Simulation Framework with a 500-Million-Persona Resource

Yang Li, Yike Wang, Di Wan, Chujun He, Zongsi Xu, Yitao Yu, Shuai Zhong, Minghao Pan 等 23 位作者

原文摘要

Simulating economic responses at society scale requires decision rules that respond to household circumstances and accounting rules that track their consequences. We introduce S$^2$-MAS, a multi-agent simulation framework with a resource of 500 million synthetic household personas. It combines trainable candidate-scoring policies, reusable financial scenarios and explicit state transitions. Trained with SFT + RLCD on 2.93 million examples, YN-mini-MLX achieves 74.55% accuracy and 38.37% question-macro F1 on held-out consumer financial-choice tasks, improving on the training-question majority baseline by 2.78 and 6.21 percentage points. A completed application evaluates 12 million distinct households in two purchase-funding conditions: 24 million one-shot choices feed 288 million monthly ledger states in 3 h 59 min on 14 GPUs. Jointly increasing purchase goals and approved limits produces heterogeneous consumption, cash and repayment paths, with much of the consumption variation arising from transitions between deferral and purchase. In a separate 1,060-household retrospective consumption experiment, standalone YN-master-9B has 3.42% lower annual MAE than the deterministic fully rational benchmark, alongside a 19.7% lower predicted population mean than observed. Information-timing tests show how year labels and later outcome information change decisions. Together, these experiments establish an executable alternative for studying heterogeneous household behavior with learned policies and explicit financial accounting.

cs.MA · cs.AI

版本更新 · v3arXiv:2603.01563

Efficient Token-Level Policy Optimization for Masked Diffusion Models via Velocity Alignment

Chenxing Wei, Jiazhen Kang, Hong Wang, Jianqing Zhang, Hao Jiang, Xiaolong Xu, Ningyuan Sun, Ying He 等 11 位作者

原文摘要

Reinforcement Learning with Verifiable Rewards (RLVR) successfully aligns autoregressive models in rigorous domains like mathematical reasoning and code generation. However, applying RLVR to Diffusion Large Language Models (dLLMs) is hindered by intractable sequence-level likelihoods, forcing existing methods to rely on unstable, high-variance approximations. We propose Sequence Likelihood-Free Policy Optimization (LFPO), an efficient framework that circumvents this bottleneck by formulating policy alignment as discrete velocity alignment. Inspired by continuous flow matching, LFPO eschews trajectory-level estimation and directly optimizes denoising logits via contrastive token-level updates, providing a stable, low-variance surrogate. Furthermore, predicting terminal solutions from intermediate states implicitly straightens the generation trajectory, enabling high-quality sampling with significantly fewer denoising iterations. Extensive experiments on reasoning and coding benchmarks demonstrate that LFPO matches or exceeds the accuracy of likelihood-based baselines while substantially accelerating inference.

cs.LG · cs.AI

新论文 · v1arXiv:2610.36462

ClusterSplat: Semantic Cluster Selection for 3D Visual Grounding in Gaussian Splatting

Jaesung Lee, Hyeontaek Hwang, Gabriel Manalu, Daeyoung Kim

原文摘要

We study 3D visual grounding in 3D Gaussian Splatting (3DGS), where a referring expression should identify a target object as both a rendered 2D mask and a set of explicit 3D Gaussians. Prior methods typically extract 3D targets by applying heuristic thresholding or top-ratio filtering to dense primitive-level relevance scores. However, these scores often vary substantially across scenes and prompts, making such heuristics sensitive and leading to unstable targeting. ClusterSplat addresses this by reframing 3D target selection as cluster-level selection over scene-adaptive candidates. The method first learns instance-aware Gaussian features, forms feature-aware seed voxels, and merges neighboring seed voxels when a description-length criterion decreases, producing a scene-specific set of cluster candidates. Given a referring expression, a query-conditioned cluster scorer ranks the cluster candidates with a lightweight MLP, and the selected cluster directly defines the explicit 3D target while its cluster score is rendered for the 2D mask. On ScanRefer, ClusterSplat achieves the best rendered 2D segmentation and explicit 3D target metrics among the compared 3DGS baselines. It also obtains the best average Ref-LERF scores, supporting the same trend under fine-grained referring descriptions. On ScanNet v2 class name queries, ClusterSplat achieves the strongest rendered 2D results and the highest 3D matching accuracy, while maintaining competitive 3D IoU. Our project page is available at https://jaesung8.github.io/clustersplat/ .

cs.CV

版本更新 · v4arXiv:2509.23166

Test-Time Policy Adaptation for Enhanced Multi-Turn Interactions with LLMs

Chenxing Wei, Hong Wang, Ying He, Fei Yu, Yao Shu

原文摘要

Large Language Models (LLMs) employ multi-turn interaction as a fundamental paradigm for completing complex tasks. However, their performance often degrades in extended interactions, as they are typically trained on static, single-turn data, which hinders their ability to adapt to real-time user feedback. To address this limitation, we first propose a new paradigm: Test-Time Policy Adaptation for Multi-Turn Interactions (T2PAM), which utilizes user feedback from the ongoing interaction as a reward signal to estimate a latent optimal policy aligned with user preferences, then updates a small subset of parameters to steer the model toward this policy, ultimately enabling efficient in-conversation self-correction. We then introduce Optimum-Referenced One-Step Adaptation (ROSA), a lightweight algorithm that operationalizes T2PAM. ROSA guides the model parameters toward a theoretical optimal policy in a single, efficient update step, avoiding costly iterative gradient-based optimization and minimizing computational overhead. We provide a rigorous theoretical analysis guaranteeing that the policy of ROSA converges to the preference of user as the number of interactions increases. Extensive experiments on challenging benchmark demonstrate that ROSA achieves significant improvements in both task effectiveness and efficiency.

cs.CL

版本更新 · v2arXiv:2609.37221

OptiCom : A Unified Framework for State-Conditioned Composition in LLM-Driven Optimization

Chenxing Wei, Sichen Liu, Lizhao Liu, Ningyuan Sun, Chen Bingzhou, Ying He, Bo Jiang, Fei Yu 等 9 位作者

原文摘要

Large language models (LLMs) are increasingly deployed to solve complex scientific and practical problems via iterative optimization. However, dynamically coordinating diverse search mechanisms as candidate quality, failure modes, and resource budgets evolve remains a critical open challenge. Targeted empirical diagnostics reveal that mechanism effectiveness is highly state-dependent. Motivated by this, we analyze how individual decisions drive final outcomes, decomposing the expected terminal improvement under a shared budget into cumulative decision opportunities minus cumulative selection losses. Guided by this opportunity-loss theoretical foundation, we propose OptiCom, a unified framework that represents LLM-driven optimizers within a shared configuration space: C=(A,Q,O,E,M,S), corresponding to artifact, query, operator, evaluation, memory, and strategy. Operating within this space, a fast LLM-based Optimization Controller dynamically composes immediate mechanisms through structured Action Packages, while a slower Strategy Adapter refines long-term selection preferences, operator weights, and templates based on accumulated trajectory feedback. Comprehensive evaluations across 32 benchmark groups demonstrate the superiority of framework: OptiCom achieves an average Max-score rank of 1.72 among 14 evaluated configurations, securing the top score in 23 groups. Ultimately, these results highlight the broad applicability and high extensibility of OptiCom as a general-purpose paradigm for robust LLM test-time scaling.

cs.AI

新论文 · v1arXiv:2610.34283HF 社区推荐

ChronoReplay: One Answer Interface Is Not Enough for Streaming Spatial Intelligence

Yihao Hu, Yitong Lin, Pan Wang, Zhihao Wen, Hang Wang, Yiwen Zhao, Deyuan Liu, Yutong Li 等 11 位作者

原文摘要

Training a spatial VLM to answer which side an object lies on does not require it to preserve the metric displacement that determines that answer, while supervision for a distance estimate may leave categorical relations largely unconstrained. When post-training emphasizes one readout, improvements can therefore concentrate on that interface while transfer to others remains weak or even degrades; we call this observed specialization interface lock-in. ChronoReplay addresses it by replaying heterogeneous spatial supervision as chronologically ordered episodes, retaining source-native targets, and balancing supervision across families. Under the same observation access, a single ChronoReplay checkpoint improves OVO-S-Bench, OST-Bench, and UCS-Bench together, including a 13.41-point gain on OST-Bench, and achieves the highest three-benchmark mean among the supervision settings compared here. Same-data controls further separate the ingredients of this transfer: answer serialization redistributes gains across OST and UCS, family-balanced learning improves over ordinary joint SFT and natural turn weighting, and chronological replay exceeds a shifted-order control by about 2.41 OST points across two paired seeds. Adaptive frame selection then changes evidence access while keeping the trained model fixed, providing an additional inference-time gain at lower visual cost. Streaming spatial performance depends on what distinctions supervision preserves, how heterogeneous histories are organized during learning, and what causal evidence inference makes available at test time.

cs.CV

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1235ACL Anthology · ACL 2026

Stability Implies Redundancy: Delta Attention Selective Halting for Efficient Long-Context Prefilling

Yujie Chen, Tailai Chen, Yifeng Gao, Zoe Wanying He, Yijue Xu Carnegie Mellon University, Shaobo Wang Qwen Team, Alibaba Group, Linfeng Zhang Shanghai Jiaotong University

原文摘要

Prefilling computational costs pose a significant bottleneck for Large Language Models (LLMs) and Large Multimodal Models (LMMs) in long-context settings. While token pruning reduces sequence length, prior methods rely on heuristics that break compatibility with hardware-efficient kernels like FlashAttention. In this work, we observe that tokens evolve toward semantic fixing points, making further processing redundant. To this end, we introduce Delta Attention Selective Halting (DASH), a training-free policy that monitors the layer-wise update dynamics of the self-attention mechanism to selectively halt stabilized tokens. Extensive evaluation confirms that DASH generalizes across language and vision benchmarks, delivering significant prefill speedups while preserving model accuracy and hardware efficiency. Code will be released at https://github.com/verach3n/DASH.git .

cs.CL · cs.CV

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1339ACL Anthology · ACL 2026

Why Are We Moral? An LLM-based Agent Simulation Approach to the Study of Moral Evolution

Zhou Ziheng, Huacong Tang University of California, Los Angeles, Mingjie Bi Beijing Institute for General Artificial Intelligence, Wanying He Wuhan University, Fang Sun, Yizhou Sun University of California, Los Angeles and Amazon, Ying Nian Wu, Demetri Terzopoulos University of California, Los Angeles 等 10 位作者

原文摘要

The evolution of morality presents a puzzle: natural selection should favor self-interest, yet humans developed moral systems promoting altruism. Traditional approaches must abstract away cognitive processes, leaving open how cognitive factors shape moral evolution. We introduce an LLM-based agent simulation framework that brings cognitive realism to this question: agents with varying moral dispositions perceive, remember, reason, and decide in a simulated prehistoric hunter-gatherer society. This enables us to manipulate factors that traditional models cannot represent—such as moral type observability and communication bandwidth—and to discover emergent cognitive mechanisms from agent interactions. Across 20 runs spanning four settings, we find that cooperation and mutual help are the central driver of evolutionary survival, with universal and reciprocal morality exhibiting the most stable outcomes across conditions while selfishness is strongly disfavoured. Beyond cooperation itself, we further identify cognition as a central mediator—most clearly through a cost of moral judgment that shifts the winning moral type across settings, with a self-purging effect among selfish agents as an additional cognitive pattern. We validate robustness across multiple LLM backbones, architecture ablations, and prompt sensitivity analyses. This work establishes LLM-based simulation as a powerful new paradigm to complement traditional research in evolutionary biology and anthropology, opening new avenues for investigating the complexities of moral and social evolution.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1424ACL Anthology · ACL 2026

The “Knowledge–Behavior Gap” in Cultural Taboo Safety of Large Language Models

Ying He, Sihang Jiang Fudan University, Xingzhou Chen, Zhouhong Gu Xiaohongshu, Yiwei Gu, Minggui HE Huawei Technologies Ltd., Shimin Tao, Mahongxia 等 9 位作者

原文摘要

Cultural taboo safety is essential for deploying large language models (LLMs), as culturally insensitive outputs may cause offense or even social harm. However, existing cultural benchmarks primarily assess cultural knowledge or values biases, while overlooking whether LLMs can recognize and respect cultural taboos, especially when taboos are implicitly hidden in seemingly harmless questions. Besides, cultural taboos are implicit, and context-dependent, thus poss unique challenges for reliable evaluation. To address these gaps, we introduce CulShield, the first public benchmark dedicated to evaluating and improving the cultural taboo safety of LLMs. CulShield spans 77 countries and regions, and includes over 2,020 taboos. It evaluates models along both explicit knowledge and implicit behaviors.Experiments on several advanced LLMs (e.g., GPT-4o-mini, Gemini-2.5-pro) reveal a clear “knowledge-behavior gap”: models often fail to apply known taboos during interaction. We further show that variations in linguistic context can significantly affect LLMs’ cultural taboo safety. Code and data is accessible here: https://anonymous.4open.science/r/CulShield-7A0E.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.2040ACL Anthology · ACL 2026

MTA:A Merge-then-Adapt Framework for Personalized Large Language Models

Xiaopeng Li, Yuanjin Zheng, Wanyu Wang, Wenlin Zhang City University of Hong Kong, Pengyue Jia City University of Hong Kong, Yingyi Zhang, Haiying He, Mengyang Ma 等 12 位作者

原文摘要

Personalized Large Language Models (PLLMs) aim to align model outputs with individual user preferences, a crucial capability for user-centric applications. However, the prevalent approach of fine-tuning a separate module for each user faces two major limitations: (1) storage costs scale linearly with the number of users, rendering the method unscalable; and (2) fine-tuning a static model from scratch often yields suboptimal performance for users with sparse data. To address these challenges, we propose MTA, a Merge-then-Adapt framework for PLLMs. MTA comprises three key stages. First, we construct a shared Meta-LoRA Bank by selecting anchor users and pre-training meta-personalization traits within meta-LoRA modules. Second, to ensure scalability and enable dynamic personalization combination beyond static models, we introduce an Adaptive LoRA Fusion stage. This stage retrieves and dynamically merges the most relevant anchor meta-LoRAs to synthesize a user-specific one, thereby eliminating the need for user-specific storage and supporting more flexible personalization. Third, we propose a LoRA Stacking for Few-Shot Personalization stage, which applies an additional ultra-low-rank, lightweight LoRA module on top of the merged LoRA. Fine-tuning this module enables effective personalization under few-shot settings. Extensive experiments on the LaMP benchmark demonstrate that our approach outperforms existing SOTA methods across multiple tasks. Our code is also available.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.421ACL Anthology · ACL 2026

CoSToM: Causal-oriented Steering for Intrinsic Theory-of-Mind Alignment in Large Language Models

Mengfan Li Huazhong University of Science and Technology, Xuanhua Shi Huazhong University of Science and Technology, Yang Deng Singapore Management University

原文摘要

Theory of Mind (ToM), the ability to attribute mental states to others, is a hallmark of social intelligence. While large language models (LLMs) demonstrate promising performance on standard ToM benchmarks, we observe that they often fail to generalize to complex task-specific scenarios, relying heavily on prompt scaffolding to mimic reasoning. The critical misalignment between the internal knowledge and external behavior raises a fundamental question: Do LLMs truly possess intrinsic cognition, and can they externalize this internal knowledge into stable, high-quality behaviors? To answer this, we introduce CoSToM (Causal-oriented Steering for ToM alignment), a framework that transitions from mechanistic interpretation to active intervention. First, we employ causal tracing to map the internal distribution of ToM features, empirically uncovering the internal layers’ characteristics in encoding fundamental ToM semantics. Building on this insight, we implement a lightweight alignment framework via targeted activation steering within these ToM-critical layers. Experiments demonstrate that CoSToM significantly enhances human-like social reasoning capabilities and downstream dialogue quality.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.715ACL Anthology · ACL 2026

Retrieval Heads are Dynamic

Yuping Lin Michigan State University, Zitao Li Zoom, Yue Xing Michigan State University, Pengfei He Michigan State University, Yingqian Cui Michigan State University, Yaliang Li Alibaba Group, Bolin Ding Alibaba Group, Jingren Zhou Alibaba Group 等 9 位作者

原文摘要

Recent studies have identified “retrieval heads” in Large Language Models (LLMs) responsible for extracting information from input contexts. However, prior works largely rely on static statistics aggregated across datasets, identifying heads that perform retrieval on average. This perspective overlooks the fine-grained temporal dynamics of autoregressive generation. In this paper, we investigate retrieval heads from a dynamic perspective. Through extensive analysis, we establish three core claims: (1) Dynamism: Retrieval heads vary dynamically across timesteps; (2) Irreplaceability: Dynamic retrieval heads are specific at each timestep and cannot be effectively replaced by static retrieval heads; and (3) Correlation: The model’s hidden state encodes a predictive signal for future retrieval head patterns, indicating an internal planning mechanism. We validate these findings on the Needle-in-a-Haystack task and a multi-hop QA task, and quantify the differences on the utility of dynamic and static retrieval heads in a Dynamic Retrieval-Augmented Generation framework. Our study provides new insights into the internal mechanisms of LLMs.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1CVF · CVPR 2026:cvf:0aee820cb9da40ff1aa294f5ba572ca5CVF · CVPR 2026

From 2D Alignment to 3D Plausibility: Unifying Heterogeneous 2D Priors and Penetration-Free Diffusion for Occlusion-Robust Two-Hand Reconstruction

Han, Gaoge, Cheng, Yongkang, Chen, Zhe, Huang, Shaoli, Liu, Tongliang

原文摘要

Two-hand reconstruction from monocular images is hampered by complex poses and severe occlusions, which often cause interaction misalignment and two-hand penetration. We address this by decoupling the problem into 2D structural alignment and 3D spatial interaction alignment, each handled by a tailored component. For 2D alignment, we pioneer the attempt to unify heterogeneous structural priors (keypoints, segmentation, and depth) from vision foundation models as complementary structured guidance for two-hand recovery. Instead of extracting priors prediction as explicit inputs, we propose a fusion-alignment encoder that absorbs their structural knowledge implicitly, achieving foundation-level guidance without foundation-level cost. For 3D spatial alignment, we propose a two-hand penetration-free diffusion model that learns a generative mapping from interpenetrated poses to realistic, collision-free configurations. Guided by collision gradients during denoising, the model converges toward the manifold of valid two-hand interactions, preserving geometric and kinematic coherence. This generative formulation approach enables physically credible reconstructions even under occlusion or ambiguous visual input. Extensive experiments on InterHand2.6M and HIC show state-of-the-art or leading performance in interaction alignment and penetration suppression. Project: https://gaogehan.github.io/A2P/

cs.CV

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition · 2026 · 正式发表

首次观测 · 记录 #1CVF · CVPR 2026:cvf:cb12f40b152f45c35ecb7e4323862b84CVF · CVPR 2026

LaMoGen: Language to Motion Generation Through LLM-Guided Symbolic Inference

Jiang, Junkun, Au, Ho Yin, Xiang, Jingyu, Chen, Jie

原文摘要

Human motion is highly expressive and naturally aligned with language, yet prevailing methods relying heavily on joint text-motion embeddings struggle to synthesize temporally accurate, detailed motions and often lack explainability. To address these limitations, we introduce LabanLite, a motion representation developed by adapting and extending the Labanotation system. Unlike black-box text-motion embeddings, LabanLite encodes each atomic body-part action (e.g., a single left-foot step) as a discrete Laban symbol paired with a textual template. This abstraction decomposes complex motions into interpretable symbol sequences and body-part instructions, establishing a symbolic link between high-level language and low-level motion trajectories. Building on LabanLite, we present LaMoGen, a Text-to-LabanLite-to-Motion Generation framework that enables large language models (LLMs) to compose motion sequences through symbolic reasoning. The LLM interprets motion patterns, relates them to textual descriptions, and recombines symbols into executable plans, producing motions that are both interpretable and linguistically grounded. To support rigorous evaluation, we introduce a Labanotation-based benchmark with structured description-motion pairs and three metrics that jointly measure text-motion alignment across symbolic, temporal, and harmony dimensions. Experiments demonstrate that LaMoGen establishes a new baseline for both interpretability and controllability, outperforming prior methods on our benchmark and two public datasets. These results highlight the advantages of symbolic reasoning and agent-based design for language-driven motion synthesis.

cs.CV · cs.CL

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition · 2026 · 正式发表