让下一篇重要论文，不再错过。

Local decoding usually fixes one weight precision for the whole generation. We study a different allocation: store 8-bit weights, decode with their top 4 bits, and redo a step at 8 bits only when the low-precision output distribution looks uncertain (entropy above a threshold or a small top-two margin). Quantization is simulated, the model is Qwen3-0.6B, and the benchmark is GSM8K. Cost is counted in weight bytes read per token. On 36724 replayed steps the entropy of the low pass ranks argmax flips with AUC 0.93. At a 15% escalation budget it catches 78% of flips and the margin gate 94%, against the 15% expected for a random gate of equal rate (a computed expectation, not a run). In 150 live problems all-high scores 64.0% and all-low 55.3% (McNemar p=0.053). An entropy gate with a 16% escalation rate scores 62.7% at 0.87 of the all-high weight bytes. Its paired difference is -1.3 (-8.0 to +5.3) points to all-high and +7.3 (+0.0 to +14.7) points to all-low (p=0.080), so the live data do not separate the gate from either end. Verified speculation reaches zero KL at similar bytes in the same cost model. A live random control, a thinking-mode phase rule, a second model and live KL were not run.

Local language models can save memory by storing weights in fewer bits, and they can reduce sensitivity to the order of answer options by averaging a few reorderings. We ask whether the first makes the second worse, and which is the cheaper way to spend a fixed budget. On Qwen3-0.6B and Qwen3-1.7B with simulated round-to-nearest weight quantization (8, 6, 4 and 3 bits, fp16 reference), read from the logits of the letters A to D on ARC-Easy, ARC-Challenge, MMLU and HellaSwag (300 items each), 6 bits lose no accuracy against fp16 (they score +1.8 (+1.0 to +2.5) and +0.9 (+0.3 to +1.5) points higher, which we do not explain), and 8 bits change it by -0.7 (-1.1 to -0.3) and -0.4 (-0.7 to -0.2) points. At 4 bits the answer-slot skew rises from 0.15 to 0.53 (0.6B) and from 0.07 to 0.12 (1.7B); Gaussian noise at the same weight SNR gives 0.15 and 0.37 (two draws) and 0.19 (one draw), so the control is inconclusive on whether rounding adds slot skew beyond equal-power noise. Flip rates rise but stay below an accuracy-matched baseline of independent uniform errors. Averaging four cyclic orders recovers +12.6 points at 4 bits for 0.6B (+5.3 at fp16). We price bits and passes in one machine-independent unit, weight bytes read times passes. In cost-matched pairs, extra orders beat extra bits only when the weights have 6 or more bits, and the decision table reduces to: use 6-bit weights and spend the rest of the budget on up to four orders, one pass each. Everything is simulated on two small models with the simplest quantizer; a title-level search found no earlier study of quantization together with option order.

Concept-based explanations in neural models are important for the development, certification, and adoption of autonomous systems. Yet, interpretability remains correlational. Without methods to guide and quantify causal relationships between explanations, a concept explanation can be provided without being necessary or sufficient for the observed output thereby appearing correct while misleading. We introduce Causal Concept Wrapper Network (CCW-Net), the first use of interventional mediation analysis from fields such as economics, medicine, and epidemiology as a training objective to align causal effects of concept-based explanations with a model's output. We further apply mediation analysis as an evaluation tool to quantify, per concept label combination, how much of any model's output is causally attributable to each concept explanation. As a training objective, CCW-Net shapes a model's concept representations such that they are both necessary and sufficient for the outputs they are credited with and independent of other concepts' contributions. Counterfactual concept representations anchor a set of causal alignment criteria: necessity, sufficiency, and independence. Necessity drives irrelevant concept effects toward zero, sufficiency ensures relevant concept representations fully reconstruct a model's output, and independence enforces the total effect decomposes linearly across concepts. We evaluate CCW-Net across aircraft control, driving, and image classification. CCW-Net provides a step toward concept-based explanations that are not merely interpretable, but quantifiably causal, supporting safer evaluation, certification, and deployment of transparent and trustworthy neural models.

Reflectance confocal microscopy (RCM) provides noninvasive, cellular-resolution "optical biopsies" of human skin in vivo by acquiring en-face images at successive depths, forming a sparse z-stack. Due to optical limitations, these stacks are anisotropic 3D volumes with lateral resolution (0.5 $μ$m) $\sim$6 times higher compared to axial resolution, which is defined by the optical sectioning (3 $μ$m), limiting the interpretation of tissue. Our goal is to provide continuous-depth visualization by interpolating intermediate sections and making the 3D volume isotropic. Such a representation permits arbitrary-direction sectioning, including histopathology-like cross-sectional examination, without requiring per-patient optimization. To that end, we introduce the first RCM-specific novel-view synthesis (NVS) approach, CD-RCM, a feedforward model that predicts realistic, unseen depths from sparsely sampled RCM stacks. Classical neural rendering methods focus on reconstruction from surface-level multi-view observations. In contrast to surface-level camera views, RCM can acquire optically sectioned en-face images of tissue beyond the surface up to 200 $μ$m. However, during visualization of the RCM stacks, observations of the shallower sections (towards the surface) obscure the deeper ones. This unique axial imaging geometry and layer-dependent anatomical organization motivated our development of a tailored architectural and training framework that explicitly accounts for RCM's depth-resolved, occlusive imaging physics. Experiments demonstrate that CD-RCM achieves high-fidelity novel-view synthesis with sub-second inference time.

South Asian monsoon precipitation depends on the thermodynamic structure of the lower troposphere, but how its ingredients combine to shape precipitation remains unclear, partly because this dependence is nonlocal in the vertical. We use symbolic distillation, which combines neural networks and symbolic regression, to discover interpretable precipitation equations from 21 years of June-August reanalysis data. Neural networks first learn the vertical thermodynamic structure most relevant to precipitation, condensing each variable into a single interpretable feature not tied to a particular vertical grid. Then, symbolic regression derives closed-form equations from these features and the surface state. Our best equation explains 1.5 times as much precipitation variance as a widely used buoyancy-based diagnostic and recovers about 85% of neural network skill while using about a thousand times fewer parameters. Unlike the buoyancy diagnostic, which adds moisture and instability, the equations treat them as competing controls, with the heaviest precipitation occurring when the lower free troposphere is humid. Over ocean, sensible heating is associated with additional precipitation mainly when near-surface air is warm and moist, improving predicted precipitation and its diurnal cycle. Although the learned features were optimized only for precipitation prediction, the equations largely respond to each feature as expected from convective physics, consistent with the reanalysis. These responses reduce the dependence of monsoon precipitation on the entire vertical column to a few simple, testable constraints that could guide data-driven precipitation models to make them more physically consistent.

Cloud-based online services face significant sub-second load fluctuations while needing to meet strict Service Level Objectives (SLOs). Cluster operators often over-provision resources to protect SLOs, sacrificing utilization and cost efficiency. Existing reactive and proactive autoscalers, serverless (FaaS) deployments, and VM/FaaS hybrid systems fail to reconcile strict SLO compliance with low cost and high utilization under fine-grained load fluctuation. We introduce Spandana, an architecture that addresses this trade off by decoupling SLO enforcement from cost optimization. A lightweight controller colocated with each application VM enforces SLOs by steering each arriving request between the VM and FaaS. Requests that can meet the SLO stay on the VM; the remaining requests are forwarded to a stock FaaS layer such as AWS Lambda. For cost optimization, Spandana's resource allocator determines the most-efficient VM provisioning by accounting for VM cost, FaaS cost, and traffic volatility, allowing the VM pool to run at high utilization. Our evaluation shows that Spandana maintains strict SLO adherence, achieves 76-86% CPU utilization, and reduces cost by 5-44% over three SOTA baselines.

Speculative decoding accelerates LLM inference by having a draft model propose multiple tokens for the target model to verify in one pass. The speedup is governed by the expected acceptance length (EAL), yet standard training objectives such as cross-entropy, KL divergence, and total variation focus only on distribution matching or single-step overlap. Recent studies instead propose block-level training objectives, aiming to optimize EAL directly. However, these objectives remain surrogates rather than exact formulations of EAL. For example, some average later acceptance over target-distributed continuations rather than the verification branches selected by accepted tokens, introducing a covariance gap that can reverse the gradient direction. We analyze this gap and derive REAL, a real EAL training loss that uses path weights for unbiased estimation and Rao-Blackwellization for variance reduction. In a controlled loss-level comparison with a Qwen3.6-35B-A3B target and a Domino drafter, REAL outperforms WTV, VSD, DBLast, and the static LK baseline on probabilistic serving acceptance length across 9 task subsets and 5 evaluation seeds. These gains require only a change to the training loss, with no modification to the drafter architecture or inference procedure, demonstrating the practicality of real EAL training. A reference implementation of the REAL loss is available at https://github.com/King-ty/speculators/tree/real-loss.

Low-rank adaptation (LoRA) is the de facto method for parameter-efficient fine-tuning of large neural networks, achieving strong task adaptation with minimal trainable parameters and low optimization cost. Beyond this efficiency, practitioners have observed two striking properties: \textit{robustness to catastrophic forgetting} and the ability to \textit{merge independently trained adapters} into a single model that performs competitively across multiple tasks. These properties are central to continual and multi-task learning, yet remain poorly understood. In this work, we provide a theoretical explanation through a unified \textit{margin-based perspective}. We analyze LoRA in multiclass linear classification under a near-orthogonal task regime with $\ell_2$-regularization, and characterize the optimal LoRA adapter across regularization regimes. Our analysis shows that, in an intermediate regime of regularization parameters, the optimal adapter aligns with the \emph{max-margin solution} on the fine-tuning data. Building on this characterization, we derive two key consequences. First, we obtain closed-form expressions for the margins on pre-training and fine-tuning data, revealing a precise margin trade-off: the regularization parameter controls the balance between retention and adaptation. Second, we analyze adapter merging, proving that merged models achieve positive margin on each task and deriving optimal mixing coefficients that balance margins across tasks and maximize the margin over their union. These results lead to a simple, training-free merging rule, which we term \textit{margin-based LoRA merging} (MaLoRA). Experiments on modern architectures and real datasets validate our theoretical predictions, showing that MaLoRA matches or outperforms several adapter-merging baselines across a range of vision and language classification tasks.

Attribution methods for Vision Transformers (ViTs) aim to identify image regions that influence model predictions, but producing faithful and well-localized attributions remains challenging. Existing attribution methods face several limitations, with gradient-based, relevance-propagation, and attention-based methods relying on local approximations, while perturbation- or optimization-based methods intervene on inputs, tokens, or surrogates rather than internal patch representations. The key challenge is that class-relevant evidence is formed through interactions between patch tokens across layers; methods that operate only on input changes, attention weights, or backward relevance signals may therefore provide indirect proxies for patch importance rather than directly testing the predictive effect of contextualized patch representations. We propose Causal Attribution via Activation Patching (CAAP), which estimates the contribution of individual image patches to the ViT's prediction by directly intervening on internal activations rather than using learned masks or synthetic perturbation patterns. For each patch, CAAP inserts the corresponding source-image activations into a neutral target context over an intermediate range of layers and uses the resulting target-class score as the attribution signal. The resulting attribution map reflects the causal contribution of patch-associated internal representations on the model's prediction. The causal intervention serves as a principled measure of patch influence by capturing semantic evidence after initial representation formation, while avoiding late-layer global mixing that can reduce spatial specificity. Across multiple ViT backbones and standard metrics, CAAP consistently outperforms existing methods in various settings and produces more faithful and localized attributions.

Two-dimensional haptic displays, such as programmable pin arrays and vibrotactile or friction-modulating touchscreens, enable tactile communication of complex shapes in a low-profile form factor. However, many shapes of interest, including CAD models, three-dimensional graphs, and map topographies, are inherently three-dimensional. This work investigates two methods for encoding 3-D information using a 2-D tactile display: (1) a single tactile height map providing coarse but quantifiable depth cues, and (2) a series of binary layers communicating sequential slices of a 3-D model, supporting more complex geometries such as overhangs and cavities. We further compare performance across two display types--a friction-modulating display and a pin array--to examine how the level of haptic abstraction affects users' ability to decode discrete 3-D information. A user study demonstrates that participants were able to effectively complete a 3-D assembly task using both strategies, achieving high accuracy across all display types. While the pin array supported faster task completion and higher-rated--though not statistically significant--user experience than the friction-modulating display for the binary-layer task, these differences did not translate to significant differences in overall usability (i.e., reassembly accuracy success). Together, these results identify design trade-offs and inform development of future surface haptic displays that support the communication and interpretation of discrete 3-D information in tactile graphics.

Diffusion Language Models (DLMs) have emerged as effective alternatives to autoregressive models for latency-critical tasks, enabling faster token generation. Recent advances employ block diffusion with self-editing capabilities, where each block refines tokens iteratively before advancing to the next one. Yet each block is only as fast as its slowest token, and the next block cannot start before it. We suggest Blurring the Block Boundaries (B3), a novel method with which two blocks are refined in parallel within a shared forward pass. B3 calibrates itself on the fly, opening the next block only once the model is ready to start generating it. It is a training-free and hyperparameter-free acceleration method which seamlessly integrates within any self-editing block diffusion model. Evaluated on state-of-the-art DLMs across 10 diverse datasets spanning varied tasks and difficulty levels, B3 results in on par generation quality and 11.5-14.5% speedup compared to the base model, pushing the Pareto frontier for latency-quality trade-offs in DLM inference.