Science & Mathematics

Wafer-map defect classification often requires labeled data and retraining when production conditions change. We present a training-free six-class framework that combines connected-component highlighting with morphology-, location-, and exclusion-aware prompts for a frozen large multimodal model. On Mixed-WM38 (6,866 maps), multi-region prompting achieves 92.9% macro and 93.8% overall accuracy. On WM811K (9,000 maps), it achieves 79.1% and 81.3%, respectively. Eight alternative rule orders achieve 90.1-91.9% overall accuracy, compared with 93.8% for the selected order.

Agent Skills allow LLM agents to combine reusable capabilities, but checking each Skill in isolation cannot rule out policy violations when they are composed. We term such failures Conflict of Skills (CoS): artifacts or assessment signals produced for one purpose may be reused downstream without preserving the restrictions or authorization needed for their new use. Using controlled executions of publicly available Skills and matched single-Skill controls, we identify four recurring mechanisms behind CoS. Building on these findings, we introduce the Conflict of Skills Attack (CoSA), which adds locally admitted Skills to an existing pool to supply missing business operations (capability supplementation) or reshape artifact exchange, decision reuse, and target/action bindings (relationship reshaping). CoSA leaves existing Skills and runtime permissions unchanged. To evaluate this threat under autonomous Skill selection, we introduce CoSA-Bench, built from real-world Skills and controlled services, with 288 tasks across recruiting, finance, IT, and business, covering data leakage, unauthorized state modification, and unsafe execution. Across five LLMs and three agent harnesses, additions capped at 5% of each original pool raise the policy-violation rate from 35.02% to 58.45%, an increase of 23.43 percentage points. Trace analysis links compliant-to-violating transitions to cross-Skill dependencies involving the additions. Finally, we develop a lightweight harness-level guard that checks information flow and authorization before consequential actions. On OpenClaw recruiting tasks across five models, the guard reduces violations from 65.28% to 6.67%, while policy-compliant task completion on matched benign controls changes from 66.67% to 65.00%.

AI assistants are increasingly asked to reason over a user's own web traces: group chats, community threads, activity logs and revision histories. On CL-bench Life, a benchmark of such real-life context learning, frontier LLMs solve fewer than one task in five. We find that the request is not the specification: over 80% of the benchmark's rubrics share less than 60% of their content words with the final request, as for a rule set two turns earlier or every record that meets the user's criterion, and LLMs answering directly pass only 67% of these implicit rubrics against 81% of the others. We propose QERA, a Question-driven Evidence Reasoning Architecture that asks before it answers. QERA first turns the request into a requirement checklist and typed information needs, filtered by code for faithfulness to the user's words; it then answers from the whole conversation, asks the same questions again with quotes verified by code, and applies a guarded local repair. It needs no training or rubrics and makes about four LLM calls per task. On all 405 tasks with four frontier LLMs, QERA attains the highest solving rate on every LLM (21.0% on average, against 18.2% for the strongest of three inference-time baselines and 14.6% for direct answering) and leads every baseline significantly when pooled across LLMs, while using fewer tokens than PSCI and 4-5% of the tokens of Ctx2Skill. Ablations on GPT-6-Luna show that the gain is settled before the answer, by a fixed answer protocol (+3.8 rubric points) and the request-specific questions (+1.6), and that it is concentrated on implicit rubrics, with checking afterwards refining mainly numeric requirements.

Automatic modulation classification (AMC) of radio signals is a core technology for spectrum monitoring and electromagnetic situational awareness. Deep neural networks (DNNs) play an important role in this field owing to their powerful feature extraction capabilities and end-to-end learning paradigm. However, their vulnerability to adversarial attacks poses serious security challenges to practical deployment. Adversarial training, a common defense, has limitations including poor generalization across attack types and high training costs. Inspired by Just image Transformers (JiT), a diffusion generative model that directly predicts clean images, this paper proposes a JiT-based denoising defense against adversarial examples for radio signal modulation classification. During training, a signal-specific 1D Transformer denoising network is optimized via flow matching to directly predict clean signal samples. It thereby learns a mapping that projects arbitrary examples back onto the low-dimensional manifold of clean signal examples. During inference, arbitrary examples are fed into the trained network and directly projected onto this manifold. This enables denoising defense without prior knowledge of attacks. Experiments on the RML2018.01A dataset were conducted without attack information, using a single empirically selected, fixed denoising start time. Compared with PGD-based adversarial training, the proposed method achieves a relative improvement of at least 20.58% in robust accuracy under PGD attacks. It also achieves a relative improvement of 10.20% in classification accuracy on clean examples. When evaluated against attack types unseen during adversarial training, the proposed method provides stronger defense than adversarial training. In particular, at higher attack strengths, it achieves a relative improvement of at least 39.72% in robust accuracy.

Unlocking medieval archives at scale is bottlenecked by the cost of line-level diplomatic transcriptions and manual abbreviation expansion. Focusing on the registers of Pope Honorius III (1216-1227) over 5,000 letters surviving in full only as manuscript images - we show that existing, line-break-free expert transcriptions (edition drafts) suffice to train end-to-end hand-written text recognition (HTR) that jointly transcribes and expands heavily abbreviated Latin. Using forced alignment on 860 unaligned edition entries across 394 pages of Reg. Vat. 9, we automatically bootstrap 13,122 training lines without manual alignment of segmented lines. While the generic CATMuS baseline struggles on this domain (18.9% CER) and a modular space-free HTR + word-segmentation pipeline (cf. Camps et al. 2021) plateaus at 15.6% WER, our from-scratch hybrid Conformer - combining CTC with bidirectional subword attention decoders, iterative self-realignment and n-best LM rescoring - achieves 4.7% CER and 8.7% WER (8.4% WER with ROVER system combination), nearly halving errors on clause-final words (22.7% to 12.2% WER). These figures are measured on the test lines that align with the transcription (86% of all test lines for the best hybrid). Furthermore, we address a tension between paleographic fidelity and editorial normalization: training on targets that preserve scribal name initials reduces the WER against the manuscript text on the 31 affected test lines from 13.3% to 9.2%, at a small cost against the expanded transcription (WER +0.3 points, CER +0.4). Finally, extensive ablations show that none of an auxiliary diplomatic CTC head, a copy mechanism, a permutation decoder or masked-image pretraining improves on direct end-to-end supervision, while strict padding-mask propagation and per-line beam termination make inference batch-invariant.

Artificial intelligence is driving an unprecedented expansion of data centres. However, this growth is increasingly stalled by grid transmission constraints that trap new data centres in years-long connection queues. Unlike electricity, computation can be moved across the globe through fibre-optic networks, but whether this spatial transfer can bypass constrained grids and accelerate global AI data-centre deployment remains unknown. Here we develop a global framework that places future data-centre capacity in energy-abundant regions and routes computational workloads through fibre-optic networks rather than transmitting electricity. We evaluate the framework through more than 11 million hourly operation simulations at subnational resolution, integrating constraints on energy systems, water, digital connectivity, land and financing risk. Results show that across multiple data-centre development and climate-policy scenarios, leveraging global data transfer allows new data centres to be sited where these conditions are favourable, lowering construction costs by 44-47% and carbon emissions by 38%. By treating the global fibre-optic network as an energy-system asset, it avoids 34-67 GW of high-voltage transmission capacity, comparable in magnitude to Italy's national peak electricity demand. Crucially, this methodology could bring more than 95% of connection-delayed data centres online in 2030, with Canada, the Nordic countries and the Arabian Peninsula emerging as preferred locations for data-centre development. Our findings establish computational mobility as an alternative to energy-system expansion, informing the coordinated global development of energy and artificial-intelligence infrastructure.

Memory has become a central mechanism for extending large language models (LLMs) beyond fixed context windows, stateless interaction, and static parametric knowledge. This survey organizes the literature around three complementary dimensions: the life cycle of memory, the storage substrate, and the system role that memory plays in LLM-based agents, while using cognitive-science-inspired memory concepts as a lightweight conceptual motivation rather than as a separate technical contribution or implementation family. On this basis, we distinguish three mechanism families: nonparametric memory, where information is stored outside model parameters in text, graph, multimodal, or database-like structures; parametric memory, where information is stored or accessed through activations, key-value caches, learned modules, adapter-based parameterizations, or model weights; and memory for agent evolution, where stored experiences support reinforcement learning, continual learning, procedural reuse, context engineering, and agent harness design. Evaluation is treated as a separate cross-cutting component rather than as a memory mechanism. Unlike surveys that focus primarily on retrieval-augmented generation or long-context modeling, this paper emphasizes the full memory life cycle: construction, maintenance, retrieval, consolidation, forgetting, governance, and evaluation. By aligning conceptual motivations, computational memory mechanisms, benchmarks, and deployment concerns, the survey provides a structured roadmap for designing memory-augmented LLM agents that are persistent, adaptive, auditable, and safe to update over time.

Test-Time Prompt Tuning (TPT) adapts large pretrained models to unseen target distribution shifts by updating only lightweight prompt tokens. Existing TPT methods are effective, but they mainly optimize each prompt update locally. Across test-time steps, prompts updated at the current step are reused for the corresponding predictions and updates, so noisy or biased unlabeled objectives can accumulate and destabilize the prompt-update trajectory. We identify two failure modes: magnitude expansion, where updates move the prompt too far, and directional drift, where consecutive updates point in inconsistent directions. To analyze and control these failures, we propose Test-Time Prompt-Agnostic Decomposition ($\mathtt{TPD}$), which characterizes observed prompt-update trajectories from a dynamical-systems view and decomposes base prompt updates into magnitude and direction. $\mathtt{TPD}$ computes a spectral radius to measure update expansion and decomposes update directions into persistent, oscillatory, and residual components. Building on this decomposition, Adaptive Koopman Control ($\mathtt{AKC}$) regulates update magnitude by shrinking updates under expansive recent dynamics, while Hankel Update Router ($\mathtt{HUR}$) refines update direction by preserving persistent components and suppressing oscillatory and residual components. Together, $\mathtt{AKC}$ and $\mathtt{HUR}$ produce a stabilized prompt update. As a prompt-agnostic framework, $\mathtt{TPD}$ can be plugged into visual and text TPT methods without modifying the backbone, prompt architecture, or adaptation loss. Experiments on 15 datasets show that $\mathtt{TPD}$ consistently improves 10 baselines, with accuracy gains of $3$--$8\%$ for visual prompts and over $2\%$ for text prompts.

For safety of the intended functionality (SOTIF) analysis according to ISO 21448, a machine learning object detector for light detection and ranging (LiDAR) point clouds can meet a performance target, a limit on the mean per-frame share of wrong detections among accepted detections, by rejecting detections above an uncertainty threshold. A threshold selected on data of one condition carries a statistical guarantee only for data distributed like its selection data, whereas SOTIF analysis covers every triggering condition. This paper presents an uncertainty evaluation in which the threshold is transferred into a triggering condition, a test decides at a stated error probability whether the target is applicable there, not applicable or not determined, and a lower bound counts the independent units an applicable outcome requires. For PointPillars ensembles on KITTI and nuScenes, counted in recorded sequences, as the dependence between frames requires, neither dataset held enough independent units for any target limit, and no transfer had a not applicable outcome. Counted in frames, the same data gave not applicable outcomes at limits fixed in advance after transfers into far distance, into rain and between two separately trained detectors, thresholds that met the target limit after a transfer rejected most correct detections, and in rain the area under the receiver operating characteristic curve did not decrease. A performance target met under one condition therefore counts as evidence under a triggering condition only after a test there, reported with its share of accepted correct detections and its number of independent recorded sequences.

The recent incident in which OpenAI agents autonomously compromised Hugging Face's infrastructure illustrates the importance of instruction hierarchy (IH): an LLM should follow lower-priority instructions, such as user requests, only when they do not conflict with higher-priority instructions, such as system policies. IH can be effectively improved by fine-tuning LLMs on simulated examples of instruction conflicts, whose diversity is key to teaching LLMs to properly handle conflicts at test time. However, our decomposition of a task into goals and constraints reveals a conflict type -- goal-goal conflicts -- that is absent from existing IH training sets from OpenAI and NVIDIA. We show the importance of this conflict type: fine-tuning Qwen3.5-9B and Gemma4-12B-it on examples of this conflict type significantly improves IH on a wide range of real-world tasks unseen in training; adding these examples to existing IH training sets for fine-tuning also gives better IH adherence. The IH gains come with trivial mean utility changes on no-conflict tasks across eight benchmarks. Our work advocates including this new conflict type in IH training, even if these examples do not reflect typical LLM use.

Implicit misogyny often requires background knowledge to be understood. To address this challenge, we propose a Retrieval-Augmented Generation (RAG) framework for detecting and explaining implicit misogyny in English and Italian, leveraging encyclopedic, slang and feminist knowledge sources. Using the ImplicIT-Mis and SBIC+ datasets, we evaluate the contribution of each knowledge source to detection and explanation quality, investigating also whether RAG addresses known LLM failure modes in implicit misogyny. Results show that RAG improves recall on ImplicIT-Mis, but does not improve detection performance on SBIC+. However, human evaluation shows that RAG-based explanations are preferred over no-RAG explanations in both languages. Feminist literature emerges as the most consistently useful source for detection and explanation.

Large language model (LLM) scores are being studied as information for scholarly evaluation. Interpreting these scores requires evidence about agreement with human judgments and its variation across evaluation conditions. Guided by an argument-based approach to validation, we compare four LLMs on 7,083 public ICLR 2024 submissions under a standard review condition and examine ten prompt conditions using DeepSeek-V3.2. We distinguish average score divergence, paper-level absolute error, and rank correlation, then examine domain differences and adjusted manuscript-feature associations. For DeepSeek-V3.2, Area Chair guidance reduces mean rating divergence from 1.460 to 0.199 points and mean absolute error from 1.577 to 0.942, while rank correlation changes from 0.377 to 0.387. Historical references combined with downward scoring instructions yield mean divergence of 0.052 and rank correlation of 0.321. Domain differences remain after global calibration. Longer titles have positive adjusted associations with rating divergence, and formula density has negative associations across all four configurations. These findings connect score interpretation to the scoring procedure, evaluation context, and manuscript information represented in the comparison. They show why aggregate alignment, manuscript ordering, and correspondence within manuscript groups require distinct evidence when evaluating LLM-generated scholarly assessment information.

Continuous diffusion language models involve two coupled design choices: how to corrupt continuous token representations and how to invert that corruption at generation time. Recent work has explored language generation on the unit sphere through geodesic and von Mises--Fisher (vMF) paths, raising the question of which forward process to use on this geometry and how much that choice matters relative to inference. We introduce Spherical Brownian Diffusion, using Brownian motion as the canonical isotropic diffusion on the sphere. Its heat semigroup provides a principled forward process, an additive dimension-free noise clock, and an exact ancestral bridge, while a clean-token posterior learned with cross-entropy supports multiple samplers. To separate corruption from inference, we derive high-dimensional approximation rates between Brownian and moment-matched vMF corruption and use this relationship for controlled substitutions. On Sudoku, Brownian and moment-matched vMF give nearly identical results across four samplers and five budgets under otherwise identical settings. In contrast, with the corruption process fixed, sampler choice produces a 25.0-point spread at a sampler budget of 180 calls. Using the ancestral bridge, we split this gain into a memory effect from re-noising the state, present on Sudoku and GSM8K, and a token-selection effect on GSM8K and largest at small budgets. Across reasoning tasks, categorical refinement improves performance, whereas unconditional text generation benefits from stochastic updates that preserve diversity.

Synthetic supervision can bootstrap semantic parsers in new search domains, but prompt-induced diversity is useful only when generations are faithful to the structured intent. We explore this for cold-start e-commerce search: 118,098 synthetic pairs produced by 6 LLMs with 9 persona-conditioned prompt families on a 47-field UK automotive schema, against 1,214 human-annotated and 15,304 production queries. We show that the IPIP personality bundle improves parser F1 in all four matched contrasts (Holm-corrected $p

Dexterous manipulation of rigid objects has reached a level of maturity where grasp stability, control, and planning strategies are successfully applied in industry and logistics. In contrast, manipulating deformable objects (cables, plastic pieces, cloth) remains far more challenging, as rigid-object methods do not account for their additional complexity. This paper identifies semantic failure analysis as a key step toward mature deformable-object manipulation and supports this claim with proof-of-concept experiments. Unlike rigid-object manipulation, where goals are well-defined in terms of criteria such as force balance and slippage, deformable-object manipulation involves a broader variety of failure modes arising from differences in materials, geometry, and behaviour. Typical failures include buckling, plastic deformation, pressure variations, and tearing. These diverse and often hard-to-detect events require distinct corrective strategies, making semantic failure analysis necessary. We define its key components and challenges and present experimental results supporting its potential as a research direction.