NeurIPS 2026 论文列表

Many systems represent graphs as sequences by visiting nodes one at a time, for example, a molecule written as a SMILES string or a program written as a token stream from an AST traversal. The same graph can be traversed in many ways, producing different sequences for the same underlying object. Sequence Transformers treat these as distinct inputs and produce distinct embeddings, even though the graph remains unchanged. Existing graph-native Transformers address this by operating directly on the graph, but many practical pipelines receive serialized inputs and cannot be redesigned. We provide a theoretical characterization in three parts. Tree-positional encoding achieves structural invariance for acyclic graphs under canonical-root serialization but is insufficient for cyclic graphs due to spanning-tree non-uniqueness. Augmenting with cycle-membership encoding derived from the canonical cycle basis introduces a traversal-invariant anchor on cycle vertices. The resulting Tree+Cycle Positional Encoding is a drop-in replacement for sinusoidal PE within graph-aware serialization pipelines (those preserving token-to-atom alignment for structural tokens), requiring no architectural changes to the sequence model. It applies to any domain where labeled graphs are given as sequences with this alignment property. We validate the approach on two testbeds. On synthetic cycle counting and ring membership tasks, it outperforms the GIN baseline by $2\times$ on cycle counting and matches it on ring membership, demonstrating that pre-computed cycle structure encoded as PE matches message-passing GNNs at end-task performance. On molecular property prediction, it achieves $0.977 \pm 0.029$ mean cosine similarity across re-rootings ($n = 24{,}960$), outperforms a parameter-matched graph-native Transformer (Graphormer) on BACE and BBBP, and enables fragment-level interpretability with $+5$ points on alert alignment over a methodologically matched sinusoidal baseline.

Effective discrete representation learning with VQ-VAE serves as a fundamental component of modern autoregressive generative models. However, it is widely known to suffer from codebook collapse and inefficient code utilization. To address this issue, we propose a simple yet effective alternative to standard vector quantization by formulating neural quantization as a joint modeling and clustering process. Specifically, we learn the latent space using a lightweight flow-matching model alongside training, corresponding to a probability flow ODE that transports an often fixed well-behaved source distribution while preserving probability mass. Based on this, we define the codebook $\mathcal{C}_1$ by transporting even-quantile partitioned prior codes $\mathcal{C}_0$ from the source space through the learned flow, and obtaining the Voronoi centroids under the induced distribution, leading a structured discretization. We evaluated on image tasks, and two tasks require high-precision tokenization, demonstrating scalability and improved expressivity of the codebook with full utilization.

High-contrast multiscale PDE problems are common in real-world applications, yet current neural PDE solvers struggle to achieve sufficient accuracy on such tasks. The Generalized Multiscale Finite Element Method (GMsFEM) addresses this by compressing fine-scale heterogeneity into localized basis functions which are used to obtain more accurate solutions than neural PDE solvers. However, constructing these basis functions requires solving many local eigenvalue problems—the major computational bottleneck. We address this issue by proposing **Transolver-GMsFEM**, a new hybrid framework that utilizes the efficiency of neural PDE solver for predicting multiscale basis functions while preserving the solution quality of GMsFEM. Experiments on 2D/3D steady-state and time-dependent high-contrast multiscale PDEs on irregular grids show proposed method achieves over $100\times$ speedup for basis construction. Crucially, our experiments demonstrate that Transolver-GMsFEM significantly outperforms state-of-the-art neural PDE solvers in both accuracy and robustness, especially in out-of-distribution tasks where pure neural PDE solvers dramatically fail.

Transmissive scenes are ubiquitous in daily life, yet reconstructing and rendering them remains highly challenging due to the inherent entanglement between near-field reflections from the surrounding environment on the transmissive surface, and the transmitted content of the scene behind it. This coupling gives rise to dual surface geometries and dual radiance components within each observation, posing ambiguities for standard methods. We present TransmissiveGS, a novel framework for disentangled reconstruction and rendering of transmissive scenes. Specifically, we model the scene with a dual-Gaussian representation and introduce a deferred shading function to jointly render the two Gaussian components. To separate reflection and transmission, we exploit the inherent multi-view inconsistency of reflections and leverage the residuals from reconstructing multi-view consistent content as cues for disentangled geometry and appearance modeling. We further propose a reflection light field that enables high-fidelity estimation of near-field reflections. During training, we introduce a high-frequency regularization to preserve fine details. We also contribute a new synthetic dataset for evaluating transmissive surface reconstruction. Experiments on both synthetic and real-world scenes demonstrate that TransmissiveGS consistently outperforms prior Gaussian Splatting-based methods in both reconstruction and rendering quality for transmissive scenes.

Public transit route planning traditionally depends on structured map infrastructure and complex routing engines, and no existing dataset supports training models to bypass this dependency. We present TransitLM, a large-scale dataset of over 13 million transit route planning records from four Chinese cities covering 120,845 stations and 13,666 lines, released as a continual pre-training corpus and benchmark data for three evaluation tasks with complementary metrics. Experiments show that an LLM trained on TransitLM produces structurally valid routes at high accuracy and implicitly grounds arbitrary GPS coordinates to appropriate stations without any explicit mapping. These results demonstrate that transit route planning can be learned entirely from data, enabling end-to-end, map-free route generation directly from origin-destination information. The dataset and benchmark are available at https://huggingface.co/datasets/GD-ML/TransitLM, with evaluation code at https://github.com/HotTricker/TransitLM.

Recent image editing systems have achieved impressive semantic understanding, visual fidelity, and instruction-following ability, while video editing remains substantially more difficult and costly. In this paper, we present a simple alternative to end-to-end video editing: instead of training a monolithic video editor, we transform a strong image editor into a video editor through anchor-based generation. Our key insight is that video editing can be decomposed into two subproblems: editing a sparse set of keyframes and propagating those edits across time. Based on this observation, we propose Anchor-based Video Editing (AVE), a two-stage framework in which a powerful image editor first performs composed editing on selected keyframes, and a motion-guided image-to-video diffusion model then generates the final video by treating the edited keyframes as fixed anchors. This design directly inherits the strengths of modern image editors while avoiding expensive end-to-end video editing training. Experiments on IVEBench and VIE-Bench show that AVE achieves strong performance in instruction following, temporal consistency, and content fidelity. Further ablations reveal that final video editing quality is strongly correlated with the quality of the image editor, suggesting that future progress in video editing may come from stronger image editing foundations and lightweight transfer to video.

Large Language Models (LLMs) have achieved remarkable progress on complex reasoning tasks, yet the mechanisms by which they acquire reasoning capabilities and the efficiency of their reasoning remain poorly understood. In this paper, we investigate the path-finding problem over directed graphs—a symbolic abstraction of multi-step reasoning. We first characterize the architectural requirements for this task, proving that one-layer transformers require $\Omega(N^2)$ size on dense graphs to implement the key DFS child-selection primitive. While a two-layer transformer implements full DFS with $O(N\log(N))$ size. Thus, a second layer is necessary and sufficient for near-linear DFS reasoning. We then extend this expressivity analysis to deeper models, proving the existence of an $L$-layer transformer that performs DFS with a look-ahead horizon of $2^{L-3}$, establishing that increasing depth yields an exponential gain in search efficiency. Finally, going beyond the expressivity, given appropriately curated training data, we show that two-layer transformers can provably learn to execute DFS via gradient flow and generalize to unseen graphs. Together, our results provide a mechanistic account of how transformers implement graph search, how depth improves reasoning efficiency, and how such reasoning algorithms can emerge through training.

Training selects for behavior, not circuitry: many weight configurations can implement the same function. Studying any single trained neural network thus risks describing accidents of one training run rather than the computation itself. This work shifts focus from what transformers happen to do to what they must do by extracting algorithmic cores, compact subspaces that are necessary and sufficient for a task and that recur across independently trained models. Here, Algorithmic Core Extraction (ACE) is introduced to isolate these subspaces, causally validate them, and recover the algorithms they implement across settings ranging from synthetic tasks to large-scale pretrained models. Markov-chain transformers embed three-dimensional cores in nearly orthogonal subspaces yet recover identical transition spectra. Modular-addition transformers form compact cyclic cores at grokking that later inflate under continued regularization, redundantly distributing the same computation across many functionally equivalent modes. This functional redundancy is found to accelerate the transition from memorization to generalization, yielding an inverse scaling law for grokking time. In six language models spanning two orders of magnitude in scale (GPT-2 Small/Medium/Large, LLaMA-3.1, Gemma-2, and Qwen2.5), subject-verb agreement is governed by a single, steerable axis that aligns across architectures. Flipping this axis inverts grammatical number throughout open-ended generation. Together these results suggest that beneath the apparent complexity of trained transformers lies a simpler, shared computational structure, and that targeting invariants rather than parameterizations may offer a more tractable path to mechanistic understanding and control.

In-context learning plays a central role in transformer-based large language models, yet its theoretical understanding remains limited. In this work, we study multiclass in-context classification under more realistic settings, including anisotropic class centers and label imbalance, by introducing a spectral data generation framework that constructs class-center matrices with a prescribed singular spectrum. We first show that the isotropy of class-center vectors, as quantified by the stable rank, improves ICL generalization performance. From a meta-learning perspective, our theorem shows that linear transformers can learn multiclass in-context classification with near-optimal per-label sample complexity, extending prior guarantees beyond the binary setting. In the test label imbalance regime, our analysis reveals that queries from majority classes are easier to classify, while those from minority classes are more error-prone; moreover, robustness to this bias improves with the stable rank. Finally, we empirically demonstrate that our theory is consistent with observations on transformers and pretrained large language models.

In image classification scenarios where both prediction and explanation efficiency are required, self-explaining models that perform both tasks in a single inference are effective. However, for users who already have prediction-only models, training a new self-explaining model from scratch imposes significant costs in terms of both labeling and computation. This study proposes a method to transfer the visual explanation capability of self-explaining Vision Transformer (ViT) models learned in a source domain to prediction-only VLM-based ViT models in a target domain via task arithmetic, without target-side explanation supervision. The proposed method endows explanation capability by adding an \emph{explainability vector} induced by explanation supervision in the source domain based on task arithmetic framework. Experiments on ten diverse target datasets show that a single explainability vector learned on ImageNet-1k augmented with patch-level explanation supervision transfers consistently, improving explanation quality while largely preserving classification accuracy. Beyond the transfer itself, we further investigate whether transfer success can be anticipated \emph{a priori} from model-internal features alone (without target-side explanation supervision), and find that a small set of such features carries non-trivial signal about the transfer outcome.

Time-parallel neural operators (FNO and successors) have transformed PDE surrogates on regular grids. Mesh-based operators (MeshGraphNets, Transolver) extend operator learning to irregular geometries but remain autoregressive in time and trained one geometry at a time, while classical parallel-in-time methods converge poorly on nonlinear advection-dominated PDEs. We propose space-time factorization: partition the model into local temporal operators (learned, graph-independent) and a spatial mixer that propagates information across nodes without graph-dependent learned parameters. This structural condition restores time-parallelism on irregular graphs and yields cross-graph zero-shot transfer once three information leaks—graph-dependent weights, graph-specific features, and graph-correlated training distributions—are controlled. The same protocol applies to two distinct spatial-mixer families tested here—fixed-spectral and learned-attention—suggesting the principle is not tied to a single mixer choice. We validate the principle in three domains. On simulating floodwave propagation in river networks, a 27K-parameter model trained on the Yangtze transfers zero-shot to the Mississippi at $R^2=0.998$, training $28\times$ faster than a recurrent message-passing baseline and scaling to $5\times$ larger graphs. On cross-geometry Navier–Stokes, both Chebyshev (SepFNO) and Physics-Attention (SepTransolver) variants reach within ~10% (relative $L_2$) of a locally-trained oracle using only 15 training geometries—to our knowledge the first demonstration of knowledge accumulation in this regime across fundamentally different topologies. On cross-city traffic forecasting, zero-shot transfer reaches within 12.9% of an oracle trained on the target city. A consistent pattern emerges across these settings: cross-geometry error decreases with both the number of training geometries and their relatedness to the test, with the rate of accumulation depending on the spatial mixer. The result opens the path for pooled-data, foundation-style models for graph dynamics on irregular topologies.

The cost of Kohn-Sham density functional theory (KS-DFT) calculations scales with the number of solver iterations, which depends on the quality of the initial guess. Machine learning methods that predict initial guesses from molecular geometry can reduce this cost, but matrix-prediction models fail when extrapolating to larger molecules, degrading rather than accelerating convergence [Liu et al., 2025]. We show that this failure is a supervision problem, not an extrapolation problem: models trained on ground-state targets fit those targets well out of distribution, yet produce initial guesses that slow convergence. Solver-Aligned Initialization Learning (SAIL) resolves this for both Hamiltonian and density matrix models by differentiating through the self-consistent field (SCF) solver end-to-end. We introduce the Effective Relative Iteration Count (ERIC), a correction to the commonly used RIC that accounts for hidden Fock-build overhead. On QM40, which contains molecules up to 4$\times$ larger than the training distribution, SAIL reduces ERIC by 37\% (PBE), 33\% (SCAN), and 28\% (B3LYP), more than doubling the previous state-of-the-art reduction on B3LYP. On QMugs molecules 10$\times$ larger than the training set, SAIL delivers a 1.35$\times$ wall-time speedup at the hybrid level of theory, extending ML SCF acceleration to large drug-like molecules.

We study transfer learning for a linear regression task using several least-squares pretrained models that can be overparameterized. We formulate the target learning task as optimization that minimizes squared errors on the target dataset with penalty on the distance of the learned model from the pretrained models. We analytically formulate the test error of the learned target model and provide the corresponding empirical evaluations. Our results elucidate when using more pretrained models can improve transfer learning. Specifically, if the pretrained models are overparameterized, using sufficiently many of them is important for beneficial transfer learning. However, the learning may be compromised by overparameterization bias of pretrained models, i.e., the minimum L2-norm solution's restriction to a small subspace spanned by the training examples in the high-dimensional parameter space. We propose a simple debiasing via multiplicative correction factor that can reduce the overparameterization bias and leverage more pretrained models to learn a target predictor.

Understanding how information propagates within vision-language models (VLMs) and large language models (LLMs) is central to interpretability and efficient deployment. Attention weights indicate where probability mass is allocated, but do not directly measure directional influence, redundancy, or functional importance. We introduce transfer entropy (TE) as an information-theoretic measure of directed information flow in Transformers, and develop tractable TE estimators for high-dimensional hidden states at the granularity of layers, tokens, and attention heads. We first analyze multimodal models, beginning with LLaVA-1.5-7B and then CLIP ViT-B/16. In LLaVA-1.5-7B, vision$\rightarrow$text TE increases with depth, suggesting that multimodal fusion is localized in the upper decoder layers. In CLIP, TE reveals late-layer redundancy in the vision tower and broader elevated TE in the text tower. We then analyze unimodal models, including RoBERTa, T5, Llama-3.2-3B, and Qwen-2.5-7B. Across these LLMs, TE reveals consistent depth-dependent structure: RoBERTa exhibits a mid-layer peak, decoder-only LLMs concentrate task-dependent computation in broad mid-depth bands, and T5 shows distinct encoder and decoder dynamics. Beyond layerwise analysis, TE also exposes token- and head-level communication roles, distinguishing sinks, broadcasters, and benign components. These signals support TE-guided layer, token, and attention-head pruning across VLMs and LLMs, consistently outperforming attention- or saliency-based heuristics under matched budgets. Overall, TE provides a unified directed-information lens for diagnosing, interpreting, and compressing Transformer-based models.

Analyzing the generalization of graph neural networks (GNNs) on node classification is challenging: message passing induces dependencies that break the i.i.d. assumption underlying standard inductive generalization theory. While distribution-free transductive theory resolves this dependency issue, existing classical bounds are either intractable or correlate poorly with empirical generalization. To fill this gap, we propose a transductive generalization bound based on optimal transport, connecting generalization to intra-class concentration and inter-class separation in the feature space. Empirically, our bound aligns strongly with the generalization gap. Building on this representation-based approach, we analyze how feature geometry evolves under repeated message passing. This explains the non-monotonic relationship between GNN depth and generalization, providing new insights into oversmoothing research. Practically, our bound serves as a geometry-aware regularizer that corrects the harmful behavior of message passing, yielding consistent performance improvements.

Optical Music Recognition (OMR), the task of transcribing sheet music into a structured textual representation, is currently bottlenecked by a lack of large-scale, annotated datasets of real scans. This forces models to rely on either few-shot transfer or synthetic training pipelines that remain overly simplistic. A secondary challenge is encoding non-uniqueness: in the popular Humdrum **kern format for transcribing music, multiple different text encodings can render into the same visual sheet music. This one-to-many mapping creates a harder learning task and introduces high uncertainty during decoding. We propose Transcoda, an OMR system built on (i) an advanced synthetic data generation pipeline, (ii) a normalization of the **kern encoding to enforce a unique normal form and (iii) grammar-based decoding to ensure the syntactic correctness of the output. This approach allows us to train a compact 59M-parameter model in just 6 hours on a single GPU that outperforms billion-parameter baselines. Transcoda achieves the best score among state of the art baselines on a newly curated benchmark of synthetically rendered scores at 18.46% OMR-NED (compared to 43.91% for the next-best system, Legato) and reduces the error rate on historical Polish scans to 63.97% OMR-NED (down from 80.16% for SMT++). We will make our model, data and generation pipeline publicly available upon publication of the paper.

Meta pseudo-labeling improves semi-supervised learning by updating a teacher according to whether its pseudo-labels help a student improve on labeled data. Existing one-step meta-gradient methods, however, differentiate through the student update, introducing mixed higher-order derivatives, extra memory cost, and noisy minibatch meta-gradients. We propose Trajectory-Matching Meta Pseudo-Labeling (TMPL), a first-order alternative that extracts labeled-data feedback from recent student optimization trajectories. We show that the one-step meta objective is locally equivalent, up to second-order terms in the student step size, to maximizing alignment between the outer gradient and the pseudo-label-induced student gradient. TMPL approximates this alignment without differentiating through the student update by matching finite differences of the outer objective and the pseudo-label-induced student objective along recent student displacements. Its implementation stores only a FIFO cache of detached scalar objective values, rather than student computation graphs or past student networks. We prove that small trajectory-matching error controls the gradient mismatch within the span of recent displacements, with an explicit residual for unexplored directions. On CIFAR10, CIFAR100, SVHN, and STL10, TMPL improves over MPL in all eight evaluated label regimes, obtains the best result in five settings, and remains competitive with strong SSL baselines. On CIFAR100 with 2500 labels, TMPL reduces peak memory by 39.7\% relative to MPL and 54.0\% relative to a raw differentiable meta-gradient implementation. Ablations identify the directional trajectory-matching term as the main source of improvement, while alignment diagnostics show that TMPL recovers MPL-like gradient alignment after a short warm-up.

Hamiltonian Neural Networks (HNNs) produce stable long-horizon rollouts by integrating a learned Hamiltonian with a symplectic scheme. Standard Monte Carlo (MC) dropout conflicts with this per-trajectory interpretation because resampling the dropout mask inside a leapfrog integrator means successive substeps follow different subnetworks, so a single MC rollout no longer corresponds to any one sampled Hamiltonian. We propose trajectory-consistent dropout, implemented as Fixed-mask MC, where each stochastic rollout selects one mask from a finite bank and reuses it across the full integration path. This restores a coherent per-rollout Hamiltonian sample and makes uncertainty decomposition operational. Crossing the mask axis with an initial-condition ensemble separates model-sample uncertainty from sensitivity to the initial state. On coupled-pendulum rollouts, Fixed-mask MC reduces mean energy drift by 72% relative to Standard MC. Under held-out mass ratios, Fixed-mask MC increases the epistemic share of predictive variance by 16.9 percentage points from in-distribution to out-of-distribution, compared with 4.2 percentage points for Standard MC, indicating a substantially stronger shift of the uncertainty budget toward model uncertainty when the physics moves outside training support. Spring-chain and transverse-field Ising experiments show that the same axis-preserving decomposition transfers to structured pairwise and latent quantum dynamics, although point-prediction and likelihood gains remain system-dependent. These results position Fixed-mask MC as a single-training-run route to trajectory-consistent, physically interpretable uncertainty in structured neural dynamical models.

Diffusion policies have emerged as a powerful policy class for offline reinforcement learning (RL) due to their ability to model complex, multimodal action distributions through iterative denoising. However, in offline RL, expressivity alone is not enough. Generated actions are supposed to remain within value-reliable regions of the action space to avoid extrapolation error. We find that standard diffusion policies often produce unstable denoising trajectories whose intermediate and final action candidates drift toward weakly supported regions, making critic-guided action selection unreliable and degrading teacher quality for one-step distillation. This issue is particularly pronounced on mixed-quality offline datasets, where suboptimal behaviors amplify trajectory instability. To address this problem, we propose Trajectory Consistency Rectification (TCR), a framework designed to promote denoising stability during both training and inference. At inference time, TCR aggregates locally consistent, high-value action candidates across denoising steps to suppress critic-unreliable outliers. During training, it adaptively reweights denoising steps to improve trajectory-level coherence. We show that this stabilization provides a more reliable teacher for one-step flow-matching distillation. Extensive experiments on D4RL and OGBench show that TCR consistently improves over diffusion-policy baselines, with significant gains in sparse-reward domains like AntMaze and Adroit. Moreover, one-step policies distilled from TCR achieve competitive performance among efficient single-step methods. These results suggest that enhancing the temporal consistency of denoising trajectories is a key ingredient for robust offline diffusion reinforcement learning.