Papers

Length generalization is the ability of sequential models to perform well on context lengths unseen during training. In this work, we show that the challenge of achieving length generalization is related not only to architectural choices such as positional encoding and the attention mechanism but also to the training procedure itself. We study how regularization affects length generalization and find that weight decay can hinder extrapolation. In contrast, dropout improves extrapolation when its placement within the architecture is reconsidered. In that regard, we show that the standard placement of dropout before layer normalization introduces a systematic distributional mismatch, and that applying dropout just before the linear projection resolves this issue. For example, a modified SmolLM3 with sliding window attention, continually pre-trained with dropout, can extrapolate perfectly to 64$\times$ on Needle-in-a-Haystack and far beyond the pre-training context size on RULER and HELMET. Mamba2 also benefits from dropout, suggesting an architecture-agnostic nature of the problem. We further propose Variance-Preserving Affine Dropout (VPAD), a new dropout strategy that substantially reduces the resulting pre-activation variance mismatch, leading to further extrapolation improvements in transformers.

Deep time-series forecasting models have rapidly diversified, yet adapting them to a specific task still requires extensive expert effort in model selection, mechanism diagnosis, architecture design, implementation, and evaluation. Existing AutoML methods are constrained by predefined search spaces, while general-purpose LLM research agents lack reliable control over experimental protocols and model promotion. We introduce EvoCast, a fully autonomous research-agent system for iterative forecasting architecture evolution. EvoCast first establishes and diagnoses a task-specific baseline through executed mechanism ablations, then generates evidence-grounded research directions from dataset characteristics, diagnostic results, prior rounds, and failure records. Its central design, cognition-authority separation, assigns open-ended hypothesis generation and code implementation to LLM agents, while deterministic program authorities control source-edit boundaries, canonical evaluation, and promotion decisions. Experimental outcomes are accumulated as evidence to guide subsequent rounds. Results show that EvoCast completes complex architecture modifications with higher implementation success and lower agent-side token/time cost, and develops task-specific architectures that outperform selected baselines, strong forecasting models, and agent baselines in three real-world forecasting cases. The code is available at https://github.com/18e0-x/EvoCast.

Credibility is a central topic for agent-based models intended to support policy-making. Simulations must not only represent the target scenarios and their core dynamics but also demonstrate that their assumptions, parameters, and outputs are empirically grounded and sufficiently accurate for their intended use. This paper addresses this challenge by presenting a general modelling framework, aligned with the Capability Approach, for building credible policy simulations that rely on data and domain-expert knowledge. It then demonstrates how it can be contextualised and implemented to study the social challenge of financial inclusion in Egypt, building on an agent-based model that represents heterogeneous individuals and firms behaving according to their financial states, barriers, opportunities, and preferences. The model is fitted to real-world data in two stages, initialisation and calibration, which respectively build representative synthetic populations and estimate behavioural parameters. By fixing the feasibility parameters, which determine agents' opportunities, and calibrating preference parameters across different population groups, we are able to distinguish and analyse the role of institutional and social barriers in the system, as well as the role of agents' motivations and priorities. This calibration stage provides transparent and group-specific hypotheses about the drivers of observed financial-inclusion gaps, which can further be analysed as gaps between agents' opportunities and realised outcomes, a very relevant insight for policy-making. This paper is thus a step towards improving the credibility and usefulness of policy simulations, strengthening the relationship between the model, the real target system, and the stakeholders who will use it. The code is available at: \url{https://www.comses.net/codebase-release/df8383cb-f49b-4f09-8cce-73603b59adcc/}.

Answer correctness is encoded as a recoverable geometric direction in the hidden states of language models. We show that the mean displacement from incorrect to correct answer representations, computed at approximately 70% of model depth from fifty labeled examples with no parameter updates, yields a scoring direction that outperforms zero-shot log-probability scoring by up to +32.0 percentage points on factual benchmarks (ARC-Challenge and MMLU) and by +38.1 to +51.8 percentage points on TruthfulQA, across five models spanning 1B to 8B parameters in three architecture families (Llama, Qwen, Gemma). The method requires one forward pass and one dot product per candidate; no generation is performed at inference. Applied as a hallucination detector on individual (question, answer) pairs, the recovered direction achieves 0.693~AUROC versus 0.578 for log-probability scoring. We additionally find that correctness directions for factual reasoning, domain knowledge, and calibrated truthfulness are near-orthogonal in representation space, revealing that language models allocate geometrically independent subspaces to qualitatively distinct notions of correct answer, with architecture-dependent variation in the degree of separation. This structure explains the observed transfer pattern—the direction calibrated on factual questions transfers within task type but not across it—and suggests that LLM calibration failures may reflect a routing problem: the model's internal representation contains more correctness signal than its output behaviour exploits.

Modal logic comprises a broad family of logics to reason about relational structures. The literature presents many such logics, displaying diverse operators, semantics, and applications. This plurality is reflected in a zoo of mechanised modal logics, which often duplicate syntax, semantics, metatheory, and reasoning infrastructure. We present a library-grade formalisation of modal logic in Lean, developed as part of CSLib (the Lean Computer Science Library) and designed around two complementary forms of reuse: vertical reuse, whereby specialised logics inherit from common abstractions, and horizontal reuse, whereby modal logic becomes a reasoning tool for independently formalised domains. Our development provides a generic framework for polyadic modal languages, reusable metatheory and proof automation, and derived interfaces for specialised modal logics. We exemplify our infrastructure by deriving basic modal logic, basic temporal logic, and Hennessy-Milner Logic; the latter subsumes and extends CSLib's previous implementation while reducing its logic-specific code by 67%. We further apply the same modal infrastructure to reasoning about mathematics (radicals of ideals), theory of programming languages (the type safety strategy for the simply typed $\lambda$-calculus), and concurrency theory (reasoning about processes in the Calculus of Communicating Systems). Our experience suggests that reuse, automation, and integration with the strong Lean ecosystem should be treated as first-class design concerns when formalising theories for shared libraries.

Diffusion models for structured scientific generation must produce samples satisfying hard geometric constraints imposed by physics, chemistry, or biology, yet inference in these settings is prohibitively slow, demanding hundreds to thousands of neural-function evaluations per sample. We unify eight state-of-the-art models spanning medical volumetrics, molecular conformations, protein backbone design, crystal structure prediction, and multi-view 3D scenes under a single abstraction, Constraint-Manifold Diffusion Models (CMDMs), in which the target distribution is supported on a manifold defined by an externally specified constraint map. All existing acceleration families fail on this class: quantization exhausts memory on high-dimensional volumetric operators; pruning breaks constraint fidelity; fast ODE solvers allow trajectories to drift off the constraint manifold; and feature-caching heuristics are blind to constraint geometry, inducing mode confusion in the high-noise regime. We introduce ManifoldCache, the first training-free, data-free accelerator designed from first principles for CMDMs. The key insight is that the conditional score decomposes orthogonally into a normal component, which enforces constraint satisfaction, and a tangential component, which navigates within the manifold. Exploiting this structure, we prove that the noise-schedule midpoint is a sharp safe-caching boundary: caching before it incurs provably bounded error, while caching after it guarantees a strictly positive fraction of trajectories suffer mode confusion, a gap that persists up to the boundary. We further prove that deeper network blocks admit provably larger certified cache strides within the safe phase, as a consequence of the score decomposition propagating through block Jacobians. The resulting schedule requires no calibration data, along with zero training overhead.

Double descent has sparked considerable interest, with recent work relating it to the data, the model and the learning configuration. Practical fine-tuning commonly involves training a small adapter on top of frozen pretrained weights, as in low-rank adaptation (LoRA). The adapter's rank is the hyperparameter that sets its capacity, yet how this rank relates to double descent has not been well explored. We quantify this relation under label noise on four vision backbones and a 7B language model with a module-matched rank sweep (MMRS), which extends past full rank and compares every rank with dense fine-tuning of the same modules, paired by seed. On DeiT-Tiny, risk is lowest at rank one and rises sharply as the adapter becomes able to fit the noisy labels, forming an interpolation cliff. Past the peak, risk falls again, but every tested post-peak rank that still saves parameters remains above dense risk. Rank-one LoRA outperforms dense fine-tuning on three of the four vision backbones, consistent with strong regularization at small rank. LoRA thus exhibits a second descent, but matches dense risk only after losing its parameter advantage, first on DeiT-Tiny at four times dense's projection weights. Code is available at https://anonymous.4open.science/r/lora-second-descent-C7B5.

Vision-language models (VLMs) are increasingly evaluated for egocentric and cross-view video reasoning, yet existing benchmarks largely focus on semantic event understanding, temporal relations, or correspondence between already observed views, leaving their ability to reason directly about future visual states underexplored. We introduce EgoExo-Next, a visual-option benchmark for dynamic visual-state reasoning, where models must identify how an observed action trajectory subsequently appears rather than predict only an action label or textual description. EgoExo-Next contains 2,503 human-curated four-choice questions from six public egocentric and ego–exo video sources and comprises four interconnected subtasks that evaluate egocentric next-state prediction, bidirectional ego–exo state correspondence, exocentric next-state prediction, and their composition in Ego-to-Exo Next-State. Extensive evaluation of proprietary, open-source, and spatial reasoning VLMs reveals a substantial human–model gap, with the best model achieving 43.81% average accuracy compared with 98.55% for humans, and the largest degradation occurring on the composed Ego-to-Exo task. These results suggest that current VLMs remain substantially limited in dynamic visual-state reasoning, particularly when temporal progression and cross-view reasoning must be composed. The benchmark is publicly available at \url{https://huggingface.co/datasets/yutongli2024/EgoExo-Next}.

LLM-agent security has produced a dense landscape of defenses - prompt hardening, content filters, permission gates, sandboxes - yet no framework tells a deployer what a defense actually guarantees, or where that guarantee comes from. We apply Verification Autonomy Levels (VAL) - L0: LLM self-declaration; L1: deterministic rules; L2: objective ground truth; L3/L4: decidable completeness; L5: impossible - to 22 agent-security defenses; the taxonomy is falsifiable (10/10 prediction hits on frozen cards, flagged). We run the first controlled deployment-value comparison: at equal budget, a VAL-guided stack (confirmation gate + schema sandbox) versus a mainstream intuition stack (prompt hardening + keyword filter), 50 scenarios, 12 attack variants, adaptive/white-box/PAIR escalation (~7,000 testbed calls; ~10,000 harness calls on AgentDojo/JADE). The VAL stack holds 0.000 attack success at 1.000 benign success (0.5% ASR at 79.7% utility on AgentDojo banking vs 4.3% undefended); the intuition stack reaches 0.000 ASR but kills all benign actions - security by model-behavior luck, not structure. Across testbeds of rising attack-surface hardness the intuition stack's zero drifts (0->1.9%->6.2%, n=16 on JADE) while the VAL stack's holds within its ODD (0->0->0), its only breach a disclosed out-of-ODD password gap (0.5%). The same zero, two different guarantees: zero is an outcome, not a guarantee.

Medical Vision-Language Models (Med-VLMs) have demonstrated strong capabilities in clinical tasks. However, they often struggle to understand anatomical structures and spatial positioning, which are crucial for medical reasoning. To address this, we propose a localization-aware enhancement to the Med-VLM pipeline, introducing improvements at three levels: data,architecture, and alignment. First, we introduce localization lens, a set of expert-validated representations that provide richer anatomical and positional context. However, as these representations increase input complexity, we integrate pixel shuffle within the model architecture to filter and refine representations, enhancing spatial information processing while preserving anatomical continuity. Lastly, to effectively align the localization lens representations with textual features, we incorporate decoupled contrastive loss (DCL) alongside the standard loss function. This ensures better feature discrimination and robustness, particularly in data limited medical settings. Through extensive evaluations on medical visual question answering (Med-VQA) datasets, we show that our methodology improves localization-driven performance across different Med-VLM architectures. Our analysis of localization-based questions further reveals that improvements in anatomy and spatial reasoning directly enhance the overall accuracy of Med-VQA upto 6.2%. The proposed approach is model-agnostic and can be seamlessly integrated into existing Med-VLM pipelines. The dataset, code, and trained models will be made publicly available at https://github.com/CVLABLUMS/localizationlens.

Adding expressive and environmental controls to a speech generator requires learning heterogeneous attributes without losing earlier capabilities. VoiceWeaver addresses this problem with structured label prefixes and staged emotion-tone-event training in a shared text-audio model. Replay-based distillation retains earlier predictions, while embedding decorrelation and attribute dropout regularize conditioning. Evaluation separates single-attribute correctness, joint emotion-event generation, and three-attribute correctness. Rounded to whole percentage points, emotion accuracy is 86% in Chinese and 83% in English. Chinese emotion-event joint accuracy is about 70%, versus 56% for mixed-task training and 48% for Ming-omni-tts; English joint accuracy is 66%. Three-attribute accuracy is 59% in Chinese and 55% in English on 500 samples each (57% pooled). Text error rates exceed those of the external TTS baselines. Audio samples are available at https://anonymous.4open.science/api/repo/VoiceWeaver1-4D88/file/index.html.

Radiology report generation (RRG) aims to convert medical images into diagnostic texts to assist in clinical decision-making and alleviate the workload of physicians. Although existing methods have made extensive progress in cross-modal interaction and the incorporation of external priors, the distribution shift of underlying representations and the undifferentiated rigid coupling of heterogeneous information cause weak visual abnormality cues to be easily diluted by massive text priors and generation inertia during decoding. To overcome this bottleneck, inspired by cognitive science, we propose a novel two-stage Homogeneous Semantic Alignment and Hierarchical Expert Routing (HSA-HER) framework. First, the model introduces an explicit homogeneous distribution constraint in the underlying latent space to effectively eliminate the cross-modal distribution shift between visual and textual features, thereby extracting purified visual features as semantic anchors that accurately align with diseases. Second, for heterogeneous clinical evidence composed of visual features, local entities, and global retrievals, we design a hierarchical expert routing mechanism guided by these disease semantic anchors. This mechanism abandons the undifferentiated rigid coupling paradigm. Specifically, it dynamically activates expert networks to perform targeted mining and semantic reconstruction on multi-source evidence, and adaptively allocates fusion weights. Extensive experiments on three mainstream benchmark datasets demonstrate that HSA-HER achieves state-of-the-art performance, accurately depicting complex imaging details and key diagnostic information.

Nowadays, emojis are often replacing words. Yet computational systems still oversimplify them. Most existing approaches treat emojis as static sentiment indicators and overlook their emotional distributions. In this study, we propose an emoji-aware emotion analysis framework based on a Twitter (X) dataset of 100,000 emoji-containing replies collected between 2020-2025. After preprocessing and text cleaning, we applied text-to-emotion classification to detect five primary emotions (Happy, Angry, Sad, Fear, and Surprise) for each message. By aggregating emotion scores across contexts in which each emoji appears, we estimate emoji-emotion association distributions and construct an emoji-emotion ranking system reflecting relative emotional dominance. Furthermore, we project emojis into the Russell valence-arousal space to enable continuous affective interpretation. Our results demonstrate that emojis exhibit probabilistic, context-sensitive emotional profiles rather than fixed sentiment polarities.

Testing deep reinforcement learning (DRL) agents in cyber-physical systems aims to uncover diverse failures before deployment, but each execution can be expensive. Surrogate-assisted testing reduces this cost by learning to predict which test configurations are likely to fail. Prior surrogates treat the system as a black box and predict pass or fail outcomes directly; we instead model how a test unfolds and estimate failure from an imagined episode. We introduce DreamTest, a world-model surrogate for testing DRL agents. DreamTest adapts a recurrent state-space model to learn agent behaviour and environment dynamics from the agent's training log. Given a candidate configuration, imagined rollouts produce a failure score that guides search without executing every candidate in a simulator or real system. We evaluate DreamTest for failure prediction, test generation, and failure diversity on Parking, Humanoid, and DonkeyCar. Mean area under the precision-recall curve (AUPRC) exceeds the strongest baseline by 97%, 12%, and 39%, respectively, and gains on five out-of-distribution test sets reach 145%, 29%, and 44%. Under the same simulator-validation budget, the best "DreamTest + search" combinations find 29%, 22%, and 79% more novel failures on average. Across clusterings with k = 2-40, failures generated with DreamTest cover the most behavioural clusters for almost all k, indicating that DreamTest consistently discovers behaviourally diverse failures.

Studying attention through the lens of interacting particle dynamics has shown how token clustering can emerge from the underlying dynamics. We extend this perspective to slot attention, a method for object-centric image segmentation and representation learning in which learned components obscure how much of the clustering behaviour is intrinsic to the attention dynamics. We therefore introduce simplified slot attention (SSA), a parameter-free variant whose dynamics are connected to soft $k$-means clustering and which provides a straightforward mechanistic explanation for the emergence of object-centric representations. On the Pascal VOC dataset, SSA achieves performance comparable to that of slot attention, demonstrating that competitive object-centric segmentation can be achieved without learned neural-network components.

As large language models (LLMs) grow in parameter count, model storage and parameter memory traffic have become major bottlenecks to efficient deployment. Codebook-based weight quantization reduces these costs, but imbalanced nearest-codeword assignments during fitting can leave some codewords insufficiently updated, limiting effective codebook utilization. To address this limitation, we propose Balanced Assignment Refinement for Quantization (BARQ), which improves quantization quality through balanced fitting of existing codebooks. Specifically, we first compute joint soft assignments between weight blocks and codewords through entropically regularized optimal transport with uniform marginals and curvature-weighted reconstruction costs, ensuring equal positive fitting mass for every codeword in the exact solution. We then refine the codewords through an assignment-weighted barycentric update, which we prove minimizes the fitting objective for fixed assignments. For finite Sinkhorn iterations, the implemented update retains this optimality provided all codeword masses exceed the denominator floor. Finally, we discard the soft assignments and use the refined codebook for standard hard nearest-codeword encoding, with our analysis establishing sufficient conditions for reducing hard-quantization distortion and evaluation loss. Across multiple LLMs, BARQ achieves lower perplexity and higher mean zero-shot accuracy than the evaluated baselines at comparable bit budgets. The code is available at https://github.com/chenhangcuisg-code/BARQ.

Understanding parameter functions helps elucidate the internal mechanisms of large language models (LLMs). However, how to connect parameters from different modules to verifiable output effects and further characterize the relationship between their functional organization and model capability remains to be explored. To this end, we introduce the downstream vocabulary lens (DV-Lens), a parameter-level interpretability framework that links native parameter directions to their downstream vocabulary responses. Specifically, we first estimate module-specific downstream Jacobians over a reference prompt set for attention query, key, value, and output (Q/K/V/O) projections and feed-forward networks (FFNs). Second, we use these mappings to project native parameter columns into the final vocabulary space, obtaining signed readouts that characterize their average local output responses. Third, we group parameter columns by their vocabulary readouts and introduce downstream vocabulary complexity (DV-Complexity), which quantifies within-group structural variation using normalized reconstruction residuals of the original weights. At the parameter level, randomized controls and finite-difference tests show that DV-Lens readouts capture non-random vocabulary structure and predict local logit changes with 98.0% coordinate-orientation agreement across 720 cases from nine models. These readouts further guide parameter ablation, steering, and swapping across 21 models, shifting target-token probabilities in the predicted directions under controlled conditions. At the model level, the joint-parameter score of DV-Complexity achieves a Spearman correlation of 0.904 with benchmark-based capability rankings across 48 language models. Together, these results provide intervention-based evidence for DV-Lens interpretations and reveal an association between DV-Complexity and model capability.

Although reinforcement learning (RL) post-training repairs the localized segmental errors of zero-shot text-to-speech (TTS), arriving at a working recipe still relies on tedious manual tuning, and whether LLM agents can take over this research pipeline is unclear. We investigate this question with AgenticTTS-Forge, a collaborative workflow that structures human guidance and agentic execution around a shared workspace, applied to CosyVoice2-0.5B. To measure what the agent automates, we audit its trajectory stage by stage against the published recipe. To measure what it exploits, we score its policies with held-out observers hidden from the agent. Our results show that the agent recovers an underspecified recipe, improves it, and, when gains stall, surveys the literature unprompted and pivots from the LM carrier to the flow carrier, halving Bad cases. However, its autonomy exposes three traps across the data, proxy, and algorithm axes: the held-out set leaks through a channel the contract never reads, a self-shaped reward inflates the proxy where it is scored, and separately tuned policies do not compose additively. These findings show that the binding constraint is measurement rather than reasoning, and can inform the design of harnesses whose contracts read every channel the agent does.

AI is not only a technological artefact; it is also a new site of political economy, legal struggle, and sovereignty contestation. AI systems depend on the continuous collection, storage, processing, and monetisation of large volumes of data. In South Africa, these datasets increasingly include behavioural, biometric, financial, consumer, health, and public-sector information. Our article uses data colonisation as a theoretical lens to analyse whether South Africa's existing legal and governance framework can regulate AI-specific risks. It argues that South Africa's current laws, including the Protection of Personal Information Act (POPIA), regulate personal information but do not adequately prevent data colonisation. POPIA remains important for privacy, lawful processing, automated decision-making, and cross-border transfers. However, data colonisation cannot solely be reduced to privacy protection. It concerns ownership, dependency, infrastructure, surveillance, regulatory power, value extraction, and the capacity of people and institutions to challenge algorithmic decisions. The article examines South African AI-related risks in various areas (e.g. cloud and digital infrastructure, biometric surveillance, and digital markets). This paper argues that South Africa's legal framework does not yet adequately regulate algorithmic opacity, externally controlled digital infrastructure, biometric surveillance, automated profiling, collective data harms, or the extraction of value from South African data. We argue for the need of a decolonial AI governance framework based on algorithmic impact assessments, public-sector AI registers, independent audits, biometric safeguards, competition-law remedies, public-interest data governance, stronger regulatory capacity, and democratic control to strengthen South Africa's digital future.

Trusted Execution Environments (TEEs) prove the integrity of code and data to a remote peer through remote attestation: the TEE issues an attestation report by signing, with a hardware-bound key, its configuration information, including measurements (cryptographic digests) of the code loaded inside it. A verifier appraises the report by matching these measurements against reference values. When the measured artifact is proprietary, however, the verifier cannot derive the reference values itself: it can only confirm that something with a given measurement is running inside the TEE—not that the program behaves as expected—forcing full trust in the reference value provider. We propose an architecture that closes this gap without disclosing the artifact, composing two TEE workloads: attestable build (Hugenroth et al., 2025), which binds a source code digest to a build artifact measurement, and attestable audit (proposed in this paper), which binds the same source code digest to the verdicts of automated audits—fuzzing, static analysis, AI code auditing, formal methods—executed inside a TEE. Both are instances of Confidential Computing Proofs: TEE attestation viewed as a hardware-backed zero-knowledge proof. Chaining the two certificates lets the verifier conclude that the running artifact was built from source code satisfying the audited properties, without disclosing either the source code or the build artifact. A proof-of-concept implementation on Intel SGX with the Gramine Library OS, running the Bandit security analyzer on Python code, demonstrates the practicality of the approach.

Partial-spoof detectors must reject synthetic content while accepting benign edits. We diagnose temporal anchors and boundary-consistency training using frozen WavLM features, nonlinguistic unit controls, and source-locked evaluation. Genuine-genuine and genuine-fake splices contrast editing false alarms with synthetic-content misses; they do not isolate a unique causal artifact. With layer-6 features, phone units have higher PartialSpoof evaluation frame EER than soft frames (11.18% versus 9.59%). In 500 retrospective PS-eval cases, augmentation reduces genuine-splice false alarms; synthetic-core misses increase at 5% PS-development FPR but not demonstrably at 1%. A 200-case Llama A follow-up retains the false-alarm reduction but does not establish a miss-rate increase. Constructed-case ranking can improve while fixed-threshold misses rise. Consistency gives no uniform gain, and encoder-layer rankings change across corpora. We report frame and event localization with three seeds, including word units, and treat PS evaluation as retrospective. Audit code, selected training scripts, and result summaries are available at https://github.com/mysxs/partial-spoof-diagnostics.

We extend the parallel integration method for mixed towers of Part~II from elementary antiderivatives to antiderivatives in special functions. The class covered is the incomplete gamma function $\Gamma(s,\cdot)$ at rational $s$, which contains $\Ei$, $\li$, $\Si$, $\Ci$, $\erf$ and the Fresnel integrals, together with the elliptic integrals $F$, $E$, $\Pi$. Each special function enters through a kernel: a known element of the tower whose antiderivative is that function. It is either fixed by residues or added as one more column of the single linear system, and no Risch differential equation is solved. We prove where kernels can have poles; the $\erf$ kernels live in the sub-critical window of radical towers. The denominator theory, degree bounds and certificates of Part~II carry over, and new criteria decide most places that Part~II leaves to a guess. Elliptic integrals are carried by the radical, and a non-torsion residue divisor becomes a third-kind term. We prove that special functions are introduced only when necessary. Elementary answers are returned unchanged, and in strict mode a special function comes with a certificate that the integrand has no elementary integral. When no complete answer is found, a partial answer with a reduced remainder is returned. All examples are computed by a SymPy implementation and verified by differentiation.

Predicting cellular responses to genetic perturbations is a central capability for virtual cells and a key step toward computational modeling of biological interventions. Most existing models directly map an unperturbed molecular profile and perturba- tion to the resulting observation without explicitly representing the latent cellular transition induced by the intervention. We introduce a mechanism-centric virtual cell world model that represents cellular state as a set of Latent Mechanism Units (LMUs) and treats genetic perturbations as actions on these latent states. Each LMU combines a reusable identity grounded in multimodal biological evidence with an observation-specific state, allowing a perturbation to induce mechanism- specific stochastic transitions before decoding the resulting transcriptional response. We train VCLMU through two-stage pretraining, first on around 200K pseudo-bulk perturbation profiles and then on gene-aligned single-cell perturbation data. Across six perturbation-disjoint benchmarks, VCLMU consistently improves perturbation- specific response recovery over strong baselines while maintaining competitive global response accuracy. We further analyze learned LMUs through enrichment between perturbation responses and LMU gene sets and show that they capture structured biological response programs. These results support mechanism-level latent state transition as a useful formulation for virtual cell models that aim to predict and interpret cellular responses to biological interventions.

Serving an LLM means setting dozens of inference-time knobs, from per-layer KV retention to per-layer expert counts. Practice sets them with mechanism-specific heuristics that return a single operating point and do not scale to layer-wise search spaces. We recast inference configuration as constrained multi-objective black-box optimization and build InferOpt, a reusable search framework that requires only variable bounds, a deterministic resource cost, and an evaluation hook. InferOpt searches on a frozen sampled proxy set, rejects over-budget candidates before any model call, tightens the budget adaptively, and re-validates Pareto representatives on full-scale dataset. One pipeline covers a 28-dimensional continuous KV space (Qwen2.5-7B) and a 26-dimensional discrete MoE space (DeepSeek-V2-Lite). On KV, post-prefill pruning cuts the 16K cache by 64.4% and TPOT by 22.9–48.5%, and the searched layer-wise budget by InferOpt beats a matched uniform budget by 7.3% and 14.0% of the Full KV reference points. On MoE, a searched top-k schedule by InferOpt removes 43.0% of routed token–expert pairs while staying within 0.59 points of the default, closer than the matched-budget baselines. Against Random Search, NSGA-II, and MOTPE, InferOpt leads on both spaces, taking the best proxy hypervolume and the lowest retention on KV and staying closest to the uncompressed reference at the lowest experts on MoE.

Aviation accident and incident investigations generate extensive unstructured textual information containing evidence relevant to the causes and contributing factors of safety occurrences. Automatically extracting such information is challenging because causal evidence may be distributed across long and complex investigation narratives. This study proposes a semantic causal-factor inference framework combining natural language processing with a variational autoencoder (VAE) to learn the relationship between aviation investigation narratives and expert-reported probable causes. Investigation narratives and their corresponding probable causes are transformed into numerical representations, after which the encoder maps narrative representations to a probabilistic latent space. The decoder estimates representations of the corresponding probable causes and is trained using an objective that combines Kullback-Leibler divergence with cosine-similarity-based semantic reconstruction. The framework was evaluated using 20,919 finalized U.S. National Transportation Safety Board investigation reports from 2005 to 2020. On the held-out test set, the predicted and expert-reported probable-cause representations achieved a mean cosine similarity of 0.786 (SD = 0.120). The predicted representations also yielded interpretable terms associated with causal information in the reports. The results demonstrate the potential of probabilistic latent representation learning for AI-assisted extraction of causal information from aviation safety narratives while retaining expert investigation as the basis for formal causal determination.

Large language models (LLMs) remain vulnerable to jailbreak attacks that conceal harmful intent within complex adversarial prompts. Existing defenses primarily rely on input perturbation or harmful-output suppression, but they rarely model where malicious intent resides, resulting in brittle protection and excessive over-refusal. We propose SENTINEL, a plug-and-play, generation-time jailbreak defense that reframes mitigation as an intent extraction problem. Our key insight is that instruction-tuned LLMs exhibit strong input–output semantic consistency: regardless of jailbreak complexity, generated outputs tend to align with the attacker's true intent. SENTINEL exploits this property by matching semantically aligned input–output regions to extract intention-revealing subsequences, scores these subsequences using refusal-direction projections to estimate harmfulness, and halts generation when necessary. Experiments on HarmBench across multiple LLMs show that SENTINEL reduces jailbreak success rates to close to 5% while maintaining low over-refusal. We further demonstrate robustness to adaptive attacks and provide a mechanistic interpretation: SENTINEL re-distributes jailbreak features from alignment blind spots to aligned regions.

Self-evolving vision-language models (VLMs), a form of self-improvement in which a model generates its own training data from unlabeled images, are a promising route toward agents that expand their reasoning capability in an unsupervised manner, without relying on ever-larger annotation budgets. Existing methods pair a Questioner that proposes problems with a Solver that answers them, but reward both roles mainly by agreement among sampled answers. Agreement is a weak proxy for truth: it cannot tell whether a question is grounded in the image, whether the proposed reference answer is right, or whether a confident majority is wrong in the same way. We present Trinity, in which one VLM plays three roles, Questioner, Solver, and Verifier, and the Verifier is a self-verifier: an exponential moving average (EMA) of the policy itself, requiring neither labels nor an external judge. The Verifier screens every generated question for image grounding and answer correctness before it becomes supervision, scores Solver reasoning against the image, and adjudicates disputes between the reference answer and a strong Solver consensus, correcting the reference and penalizing the Questioner when the consensus is right. Trained on images alone, Trinity improves Qwen3-VL-8B on mathematical visual reasoning and on science benchmarks with biology content, for example, +8.6 points on the biology split of SciVQR and +12.8 on MathVerse, and its reward dynamics behave as a healthy self-play curriculum should. These results suggest that a self-evolving multimodal agent can strengthen its scientific reasoning from unlabeled scientific images alone, with the model itself serving as the verifier.

Status updates let a receiver track a changing process, but a sender that controls when they are sent can steer the receiver's estimate without misreporting. We study competition between two such senders with opposing preferences, who report on a binary continuous-time Markov chain using state-dependent Poisson rates under individual budgets. The receiver follows the reports through a zero-order-hold estimator only if this yields at least the long-run weighted correct-estimation utility of a prior-based default estimate. We model the interaction as a Stackelberg game in which the sender favored by this default leads, the opposing sender follows, and the receiver then selects its estimator. We characterize the pure-strategy equilibria in closed form. When the follower can obtain positive utility against a silent leader, limited state-$0$ reporting by the leader merely frees part of the follower's budget for state-$1$ reports, so every optimal leader policy is either silence or full-budget state-$0$ reporting, depending on explicit budget thresholds. A single sender with the same total budget leaves the receiver at its default utility, whereas the full-budget intervention outcome makes the receiver's participation constraint slack and strictly improves its utility. At the intervention threshold, silence and full-budget intervention yield identical sender payoffs but different receiver utilities. Numerical results illustrate these regimes and compare equilibrium performance with a receiver-optimal reporting benchmark.

Large language models (LLMs) remain highly vulnerable to jailbreak attacks that induce harmful behaviors and circumvent safety alignment. To defend against such attacks, adversarial training paradigms have been proposed to first simulate failure modes and then train the model to correct them, yielding promising improvements in safety alignment. However, these methods typically construct adversarial samples either by encouraging fixed harmful target completions or by performing targeted activation ablation derived from fixed benign–harmful data pairs. As a result, the generated adversarial samples tend to induce homogeneous harmful behaviors that poorly reflect the diversity of behaviors elicited by real-world jailbreak attacks. This behavior-level narrowness fundamentally limits their robustness. To address this issue, we propose a target-free adversarial training framework that generates adversarial samples in an unsupervised manner. By amplifying and diversifying behavior-level shifts in the model's latent space, our approach produces semantically diverse adversarial samples that induce a wide range of harmful behaviors. This expanded behavioral coverage exposes more diverse failure modes and thereby improves safety alignment. To quantify this effect, we use semantic entropy as an output-level measure of adversarial behavioral diversity. Empirically, our method elicits diverse harmful behaviors in the target model, substantially mitigating behavioral narrowness and improving robustness to jailbreak attacks.

This paper addresses asymptotic tracking with prescribed transient and steady-state performance for a class of uncertain high-order nonlinear systems with uncertain dynamics. The proposed controller combines a barrier-type transformation of a filtered tracking error with a robust integral of the sign of the error (RISE) mechanism, yielding a continuous and low-complexity design. The proposed method does not employ neural, fuzzy, or other global approximation structures. It is shown that all closed-loop signals remain bounded, the prescribed performance funnel is preserved for all initial conditions in the funnel, and that the tracking error converges asymptotically to zero under appropriate gain selection. The analysis provides a continuous-feedback synthesis that integrates prescribed performance control (PPC) with the RISE methodology. Simulation results on a ship-steering example illustrate the effectiveness of the proposed approach.