Measurement exploration reveals how interacting local controllers behave but can alter the steady-state regime a network operator seeks to optimize. Exploratory rewards may therefore fail to identify a configuration's no-exploration value: under full exploration, identical logged distributions can conceal different target values. Identification requires accounting for both action-law mismatch and shifts in the steady-state context distribution. We introduce CAUSAL-UCB, which selects configurations using optimistic bounds on target causal values. Its interference-aware clipped doubly robust evaluator produces predecision certificates separating statistical variation from calibrated identification, transport, and temporal-dependence errors. For any finite horizon and predetermined schedule of measurement-exploration rates $\{\eta_t\}$, we prove lower bounds of $\Omega((1+\sum_t\eta_t)^{-1/2})$ on minimax inference error and $\Omega(\sum_t\eta_t)$ on measurement-induced deployment loss in a stationary, potentially slow-mixing family. We bound regret under valid certificates and selected-information growth, and establish an anytime finite-sample certificate with a shared radius for estimation, selection, and reporting in a restricted known-local-interference model. In a high-mismatch benchmark with 24 paired seeds and 200 rounds, CAUSAL-UCB attains target regret of $1.732$ versus $6.326$ for certified IPS-UCB, with deployment gaps of $30.607$ and $23.217$, respectively. Independent-audit coverage tests and trace-driven semi-synthetic studies further assess certificate coverage and deployment trade-offs.
In this paper, a 3D backbone network (SpconvTransformer) based on the fusion of convolutional architecture branch and Transformer architecture branch is proposed. The convolution branch uses 3D sparse convolution and submanifold convolution to learn local voxel features and semantic information, and the Transformer branch uses the self-attention mechanism to model the remote relationship between voxels, uses the broad receptive field to capture rich context information, and complementing the advantages of the two architectures by fusing the local features extracted from the two branches and the global features solves the problem of limited receptive fields of 3D convolutional neural networks. The experimental results show that the average accuracy of SpconvTransformer in 3D object detection and bird's-eye view detection tasks reaches 81.45% and 88.09%, respectively, and the average directional similarity reaches 91.10%, which is significantly better than other existing algorithms.
How does using less conversational history change reader-token use and benchmark scores? We examine this question on 1,100 BEAM questions. We held the retrieval ranking and reader constant across five fixed passage budgets, and mean assigned scores rose as the passage budget grew. The score gain per additional token became smaller at the upper tested budgets. Answerable questions generally received higher scores with more context. Pooled abstention scores sometimes rose as the passage budget shrank. We also tested message retention in a controlled task with explicit conversational-state targets and compared retained user messages with token-matched recency. Qwen matched the full state more often for direct requirements when user messages were retained. Its full-state matching declined when answers depended on references to earlier offers. Mistral produced near-floor results across the retention conditions. A scoring audit leaves the correctness of BEAM answers unresolved. These experiments characterize the token and score consequences of the tested memory budgets. Reliable budget selection depends on retaining the evidence each question requires and validating the measure used to assess answer quality.
Online 3D street-scene asset construction turns multi-view video into a continually updated scene of independently editable objects. The central challenge is to accumulate partial, intermittent observations under motion and occlusion without fragmenting identities or mixing evidence across objects. To address this challenge, we propose Video2Asset, an online framework for constructing persistent 3D assets from multi-view video. Using occupancy as an intermediate representation, we construct identity-aware observations with Historical Carrier Guidance for instance decoding and an identity encoder fusing instance-aligned 3D features with multi-view appearance. Accumulated-Evidence Association compares current observations with identity, geometric, and appearance evidence accumulated within asset records to determine identity inheritance and update world-coordinate states. We instantiate meshes from an asset library using these states and combine them with an occupancy-derived static background to support scene visualization, mesh replacement, object deletion, and trajectory editing. Experiments on several datasets demonstrate improvements in temporal identity consistency and asset quality across multiple occupancy models, including a 15.4% reduction in asset fragmentation on nuScenes with SGE.
Prompt learning adapts vision-language models such as CLIP by optimizing a short learnable prompt. The cross-entropy objective pulls this prompt toward base-class specialization, whereas generalization to unseen classes benefits from staying close to CLIP's zero-shot feature space. Because both objectives act on the same prompt, their gradients conflict at convergence, and in our experiments no single-prompt loss design clearly improves on the cosine baseline. We propose Decoupled Mode Connectivity (DMC), which gives each objective its own prompt. Starting from a base-trained prompt, DMC trains a specialization prompt and a generalization prompt, and merges their text features with a fixed coefficient at test time, at the inference cost of a single prompt. A linear mode connectivity (LMC) corridor in text-feature space keeps the classification loss low for every classifier interpolated between the two endpoints, and a Visual Anchor KL term preserves CLIP's zero-shot class-similarity distribution of each image during specialization. We further argue that such regularizers should satisfy class-permutation invariance (CPI), and show via Fano's inequality, in an idealized model, that violating CPI lower-bounds the drop in novel accuracy by a term proportional to the mutual information between the prompt and the base-class labels. Across 11 datasets, DMC improves the average harmonic mean of base and novel accuracy over MergeTune for both CoOp and KgCoOp, and its interpolation curves pass beyond the single-prompt operating point, shifting the base-novel frontier rather than trading along it. Code is available at https://github.com/IemProg/DMC.
Parameter-efficient fine-tuning (PEFT) has been increasingly adopted to fine-tune pretrained models with low computational cost and is therefore also preferred in the federated learning paradigm to reduce computational and communication overhead. In this work, we investigate the sparsely explored problem of the risk of private data leakage in LoRA-adapted Vision Transformers (ViTs), where low-rank updates weaken the gradient signal available to a server with access to individual client updates. We show that quality of reconstruction is limited by two coupled ambiguities: reduced information in LoRA gradients and missing spatial cues from frozen positional embeddings. To address this, we propose a spatially anchored generative inversion framework that incorporates server-prescribed trainable positional embeddings into ViT blocks during fine-tuning at clients. The resulting positional gradients act as layer-wise spatial anchors, providing additional positional cues, while the generative prior keeps reconstructions on a realistic image manifold, reducing duplicate-data and patch-permutation artifacts, {which produce similar gradients as those produced when the target client images are used}. Adding our proposed spatial anchoring to GIAS improves PSNR by 10.14 dB, SSIM by 0.45, and LPIPS by 0.098 on CIFAR-10, as well as PSNR gains of 7.83 and 11.74 dB on high-resolution CelebA and CARS datasets, respectively, in the FedSGD setting. Such gains in reconstructed image quality metrics are also maintained at higher batch sizes and in FedAvg settings. These results indicate that LoRA-based PEFT does not, by itself, prevent data leakage in FL, and carefully incorporated positional embeddings can help in exposing private information.
Mutation testing for deep learning (DL) faces a cost-realism trade-off. The pre-training approach produces realistic but computationally expensive mutants, whereas existing post-training operators modify trained models efficiently but produce mutants that less closely resemble real faults. This study asks whether the trade-off is necessary: can the weight-space effect of a fault-informed pre-training mutation operator be predicted from a trained model, without retraining it? We answer this question with TensorMutation, which represents a trained model as a neural graph and uses a mutation-conditioned graph neural network (GNN) to predict parameter changes. The GNN is trained offline and reused across unseen models so that TensorMutation can generate executable mutants without retraining. We construct a corpus of 25,252 original-mutant pairs from 71 multilayer perceptron (MLP) classifiers using 11 fault-informed pre-training operators, including 25 newly curated subjects. TensorMutation partially reproduces the behavioral effects of pre-training mutations on nine held-out subjects, and direct weight-space supervision significantly improves weight-space fidelity. On 12 real-fault subjects, TensorMutation achieves the highest median behavioral similarity among the evaluated approaches. The improvement over the post-training tool DeepMutation++ is statistically significant (Holm-adjusted p=0.018), whereas the differences from the pre-training tool DeepCrime and the norm-matched random control are not significant. These results show that fault-informed pre-training mutations leave learnable structure in weight space that transfers to unseen models.
Antarctic whale vocalization detection requires temporally localizing call boundaries in long underwater recordings and classifying them into predefined call types. Training reliable machine learning models for this task is difficult due to the variability and sparsity of whale calls in passive acoustic monitoring datasets. These difficulties are further amplified by time-varying noise that can mask whale calls and by site-specific ambient sound characteristics to which a trained model may not generalize. In this work, we present NAVE, a Conformer-based detector that improves whale-call detection via three components. First, its input representation combines per-bin complex mean subtraction, which suppresses constant site-specific ambient sound characteristics, with an additional per-channel energy normalization (PCEN) channel that adapts to temporal variations in the noise floor. Second, a frequency-dynamic stem processes these noise-aware representations with convolutions that adapt across the characteristic frequency bands of the target call types. Third, we use a Conformer backbone with a convolution kernel matched to the duration of the target calls that adds an inductive bias and improves temporal localization. We evaluate NAVE on BioDCASE 2026 Task 2 using leave-one-site-out cross-validation across the validation sites, which approximates deployment on unseen recording sites. A single NAVE model with a tuned per-class threshold exceeds the stronger WhaleVAD-BPN baseline, using no boundary-proposal network and no search over post-processing parameters. A controlled ablation study shows that the normalized front-end and a duration-matched convolutional kernel each contribute substantially to recovering the D downsweep, the hardest class for previous detectors, with the wide kernel giving the largest aggregate gain.
Language model deployment often requires different trade-offs between competing objectives as user preferences change, yet standard post-training methods typically produce a single policy for a fixed objective weighting. We introduce LambdaMix, a new parameter efficient method for learning a preference conditioned family of policies with group-relative online reinforcement learning. LambdaMix jointly trains one low-rank adapter per objective over a frozen base model. One preference vector selects the adapter mixture that generates responses and weights the separately normalized rewards used to update the adapters. A single training run supports changing preferences at inference without further optimization. We experiment with four models on helpfulness and harmlessness and on solving math problems while following instructions. We also evaluate the two largest models on rewriting comments to preserve wording while reducing clues about personal attributes. On the first two tasks, LambdaMix achieves higher hypervolume than all evaluated conditioned and policy merging baselines across all four models. LambdaMix uses fewer training completions and stores fewer adapter parameters than a set of GRPO specialist policies. With architecture and initialization fixed, varying adapter mixtures and objective weights together gives wider score ranges and higher coverage than varying either alone.
Model-based reinforcement learning (MBRL) approaches have demonstrated strong performance across diverse domains in recent years, learning accurate world models from agent trajectories. However, despite this progress, current approaches still struggle to surpass human performance on tasks that require long-term memory and planning. In this work, we address the problem of long-term memory in MBRL by moving beyond the commonly studied paradigm of architecture design. We introduce Test-Time Model-Based Reinforcement Learning (TT-MBRL), using gradient descent at test time to encode the memory of episode observations into world model parameters. This enables the agent to learn a distinct world model for each episode, improving memory capabilities compared to traditional approaches that solely rely on contextual memory. We first establish a proof of concept on a simplified two-dimensional variant of the Memory Maze environment, which we refer to as Memory Maze 2D, showing consistent improvements across different world model architectures. We then demonstrate the effectiveness of TT-MBRL in the 3D domain, achieving new state-of-the-art results on the Memory Maze benchmark. Finally, we show that TT-MBRL maintains strong performance on non-memory benchmarks such as DeepMind Control and Atari 100k.
Trial-and-error FE simulation of sheet forming could be computationally expensive, neural-network-based learned simulators are used to accelerate design. These simulators should accurately capture localised responses, such as thinning around sharp radii or near ribs, which determine formability. Transformer-based neural operators propagate information efficiently across the mesh but tend to smooth such localised responses. This study develops BiLoGo, a Bipartite Local-Global learned simulator with an FE-guided design that connects local node-element message passing with a Transformer-based neural operator for long-range information exchange. BiLoGo has three modules: a bidirectional node-element GraphSAGE model that learns nodal-elemental coupling on a bipartite graph and predicts each quantity on its native FE entity without interpolation; a dual-stream Transformer-based neural operator that processes blank deformation and tool contact with separate token sets, preventing localised contact information from being averaged with the whole-blank state; and a fusion and element update module that combines global and local features via gated residual fusion and updates element features conditioned on element-node kinematics. Evaluated on a dome-shaped cold forming case and two corner-shaped hot stamping cases (one with ribs), BiLoGo achieves the lowest positional and threshold-based thinning errors on all three cases compared with single-level and hierarchical graph-based models and Transformer-based neural operators. Near the ribs, the smoothing index of predicted thinning increases from 0.41-0.56 for Transformer-based neural operators to 0.94 for BiLoGo-U. Compared with the hierarchical graph network CAtt-BiUGNN, BiLoGo-T reduces mean positional error by 15.5% on the Corner case and BiLoGo-U by 8.3% on the Corner+Rib case, with 66-69% less training memory.
Dense video object captioning (DVOC) localizes, tracks, and describes object trajectories within a video. Existing methods process each camera view independently, producing separate camera-local tracklets and captions for the same person across cameras. These outputs lack a unified description and leave complementary appearance evidence across views unused. We therefore formulate cross-view DVOC (CVDVOC), which associates these tracklets into scene-level global trajectories and generates one caption for each trajectory from synchronized views. To support this task, we construct CRTrack-Cap from CRTrack by auditing text-identity links and converting usable referring expressions into canonical captions. It retains 425 global trajectories and 838 local tracklets across 13 scenes, with canonical captions for 335 identities. We adapt CHOTA to the cross-view setting as CV-CHOTA to jointly evaluate tracking and caption quality. Our reference framework combines View-Exclusive Cross-Camera Association (VECA) for grouping tracklets with Conflict-Evidence Gating (CEG) for integrating caption evidence using its reliability and disagreement across views. Against three single-view methods adapted with MvMHAT for cross-view association, our framework achieves CV-CHOTA scores of 83.43 and 33.48 on the scene-disjoint in-domain and cross-domain tests, respectively. Together, CRTrack-Cap and our framework provide a benchmark and a strong baseline for describing the same person consistently across camera views.
We present a unified potential method for contextual search and bilateral trade with one-bit feedback and obtain order-tight algorithms for both problems. Our framework starts from a Bayesian perspective in which the unknown target (for contextual search) and the unknown pair of valuations (for bilateral trade) are sampled from a known prior distribution. In that model, the decision maker bases its prediction on the posterior distribution, obtained by conditioning on feedback received in previous rounds. For Bayesian contextual search, we present an algorithm with a total loss bounded by $O(\mathrm{Tr} \sqrt \Sigma)$, where $\Sigma$ is the covariance matrix of the target's prior; and for contextual bilateral trade, our algorithm's regret is bounded by $O(\mathrm{Tr} \sqrt{\Sigma_S}+\mathrm{Tr} \sqrt{\Sigma_B})$, where $\Sigma_S$ and $\Sigma_B$ are the covariance matrices of the seller's and buyer's valuations. We then use ideas from this Bayesian setting and adapt a ``dual'' of this trace-square-root potential to the classical worst-case setting, where the target and valuations are chosen adversarially from the unit Euclidean ball. For both contextual search and bilateral trade, we obtain tight $O(\sqrt d)$-regret algorithms. These results match the existing lower bounds up to a constant factor and improve on the previous guarantees of $O(d\log d)$ for contextual search and of $O(d6^d)$ for bilateral trade.
Large Language Model training is increasingly constrained by memory capacity. Activation Checkpointing reduces activation memory by avoiding storing selected activations during the forward pass and rematerializing them during backpropagation, thereby trading memory savings for additional computation. Although this technique is essential for scaling model training, rematerialization can be prohibitively expensive, particularly for Transformer linear layers. Approximate computation could reduce this overhead, but conventional approaches either introduce errors that accumulate over the backward pass or fail to exploit information available from the initial forward pass. We introduce a novel approximate rematerialization method that can yield unbiased gradient estimates: using information collected during the initial forward pass, the method selectively recomputes only the most important layer components, either through greedy selection or randomized sketching. Ablation studies show that the resulting gradients closely match the exact gradients. When applied to feed-forward layers, attention layers, or both, our method can recompute only two-thirds of the layer components in a Llama-488M model, while increasing perplexity by just 1.7%, 1.2%, and 2.5%, respectively.
We study kernel-based learning of analytic operators from finite samples, with a focus on how the regularity of the target should determine the choice of the kernel. We establish embeddings of analytic operators into vector-valued reproducing kernel Hilbert spaces and derive finite-sample, high-probability guarantees that quantify the uncertainty due to random sampling: in the noiseless setting, suitable analytic kernels yield rates exponential in a fractional power of the sample size, while Sobolev-type kernels yield polynomial rates. We apply the theory to the solution operator of a parametric elliptic PDE, where the training data are themselves finite-element approximations, and obtain bounds that separate the sampling and discretization contributions to the error and prescribe how the mesh size should be coupled to the sample size. The results are illustrated numerically.
Auditing outsourced LLM fine-tuning requires checking an untrusted provider's training while controlling verification cost. Uniform sampling ignores training-time signals, while checking a suspicious record can require replaying several preceding steps. We present ReplayAudit, a telemetry-guided auditing system that seals training records before challenges and independently replays sampled intervals from sealed checkpoints. Loss changes and parameter-update norms guide stratified random sampling; replay determines whether the selected computation agrees with its records. We relate sampling probabilities to replay work and evaluate the system on real LoRA training. A 7B BF16 prototype rejects a false update-norm report during a random audit; targeted checks also reject false loss reports and skipped updates. We calibrate sampling budgets offline to at least a 90% probability of directly selecting an injected step. For two learning-rate deviation settings, executing these budgets on the same honest 1.5B trajectory reduces median complete-audit wall time by 49% and 67% relative to uniform sampling. When skipped updates evade the high-anomaly stratum, the corresponding audit instead costs 38% more. These results show how deviation capture and actual replay work jointly determine the cost of telemetry-guided auditing.
We study coresets for clustering problems, focusing on the classical $k$-median and $k$-means objectives. A coreset is a small weighted summary of the input data that approximately preserves the clustering cost for every choice of centers. Such summaries are a central tool for scaling clustering algorithms to large datasets. In this work, we consider the capacitated setting, where the clustering is subject not only to choosing centers but also to respecting prescribed cluster-size constraints. This model is a natural generalization of standard clustering and has been widely studied, but its coreset complexity is less well understood. For finite metric spaces, we obtain optimal-size coresets; surprisingly, these match the best-known bounds for unconstrained clustering. We also give improved coreset bounds in Euclidean space, notably the first coresets with near-linear dependence in $k$ while being independent of the dimension and input size. The main new technical ingredients are concentration bounds for the dual solutions of min-cost flow problems, which allow for both significantly simpler and tighter analyses of the underlying sampling distribution used in the coreset construction.
Clear cell renal cell carcinoma (ccRCC) is the most common kidney cancer in adults. We present a region-aware architecture for CT-based ccRCC classification. Tumor and 0--10 mm peritumoral context are pooled separately from shared frozen DINOv3 feature maps. A shared position-aware slice aggregator precedes fixed training-only principal component analysis (PCA). The resulting representations are fused with radiomics using two scalar feature gates. Evaluation on KiTS23 and 111 external cases yields internal test and external AUCs of 0.8078\pm0.0154 and 0.9094\pm0.0015. Compared with matched single-tumor readout, mean AUC increases by 1.50 and 2.34 percentage points, respectively. Early-PCA and reconstruction controls yield lower internal mean AUCs, while external AUC is similar for early and late PCA. These findings highlight the joint role of regional readout and slice aggregation in adapting frozen visual features to renal CT.
Private, personalised language models must run on local consumer devices, where large, frontier models cannot fit, and retaining long contexts is non-trivial. Sequential memory---the key-value (KV) cache---grows linearly with context length, quickly dwarfing the model itself: for Qwen3-1.7B-Base, 200,000 retained entries require 21 GiB, 6.6 times more than the model weights. Yet the KV-cache compression methods we measured fail beyond moderate compression ratios. We introduce Transformer-FSQ, addressing extreme KV-cache compression by replacing standard key and value projections with a finite scalar quantisation bottleneck. Our cache stores a packed vector of discrete integer codes per token, with no per-token quantisation metadata. Converting a pretrained transformer trains only these projections on 0.25 tokens per original parameter. We evaluate Transformer-FSQ against five baseline families on models from 0.6B to 8B parameters. On long-context retrieval (RULER), from 12$\times$ to 97.5$\times$ Transformer-FSQ is the most accurate method we evaluate, across scales. A fused Triton kernel decodes directly from the integer codes; with baselines given their authors' kernels where available, it decodes at $1.01\times$ (95% interval $0.98$--$1.05$) the throughput of native SDPA attention on the uncompressed Qwen3-8B-Base at $32\times$ and $1.54\times$ at $97.5\times$. Extreme KV-cache compression is therefore a practical way to extend how much context a fixed device can hold.
Rubric-based reinforcement learning (RL) extends RL with verifiable rewards to open-ended tasks by scoring each response against query-specific rubrics. Under group-relative optimization, however, a rubric that every sampled response fails (unreachable) or passes (saturated) gives all responses the same outcome and no learning signal; in Qwen3-4B medical training, 60.88% of query-rubric observations are inactive. We find that the two states lack different supervision: stating an unreachable rubric in the prompt raises its satisfaction from 1.13% to 91.25%, whereas responses that all pass a saturated rubric still differ in quality. We therefore propose RuACT (Rubrics as Adaptive Critics and Teachers), which assigns each inactive rubric the role that supplies what it lacks. Rubric as Teacher uses unreachable rubrics as privileged information for reward-gated on-policy self-distillation, and Rubric as Critic decomposes saturated rubrics into child rubrics that separate current responses while sharing the parent's weight. RuACT outperforms GRPO, rubric reweighting, and rubric-guided exploration on all four medical and scientific benchmarks with Qwen3-4B and achieves the best average with Qwen3-1.7B. On initially unreachable rubrics, unguided recovery rises from 35.02% to 42.86%, while child refinement improves reward-quality alignment on independent rollouts from 57.74% to 60.77%. Code is available at https://github.com/ZJU-REAL/RuACT.