arXiv:2606.02679v1 Announce Type: new
Abstract: Multimodal systems often benefit from combining information across language, sound, and visual streams, but this benefit is not guaranteed. A modality that is useful for one input may become distracting for another, and local feature responses within the same modality can disagree with evidence from other sources. This work investigates how to adjust multimodal representations before they are merged by a downstream predictor. We develop a compact calibration module that compares each modality with the others at the summary level, extracts cues of cross-source support and conflict, and converts these cues into instance-wise and dimension-wise modulation signals. The calibration is applied to the original modality features rather than to already fused representations, enabling the model to suppress misleading components, preserve weak but useful evidence, and emphasize responses that are better supported by the current multimodal context. The module is designed as a plug-in component and can be attached to different fusion backbones without changing their prediction heads. Across five benchmarks covering sentiment understanding, action recognition, audio-visual event detection, and audio-visual emotion classification, the proposed pre-combination calibration strategy improves performance under both sequence-based and convolutional fusion settings. Additional analyses under modality removal, synthetic corruption, training dynamics, and feature-level visualization show that calibrating signals before fusion can reduce interference from unreliable modalities and produce more stable multimodal optimization.
Science Journals
arXiv:2606.02631v1 Announce Type: cross
Abstract: This paper studies whether audio, images, and video can share a common wavelet token schema rather than relying on separate modality-specific latent grids. It introduces a preliminary continuous-token model built around a one-level Haar DWT/IDWT frontend, a shared coefficient-token layout, optional structural metadata, lightweight modality value adapters, and a shared token-wise encoder-decoder trunk. On Speech Commands, EuroSAT RGB, and DAVIS 2017 data, a dense shared model reaches 39.92 dB audio, 29.37 dB image, and 23.93 dB video PSNR. A matched-rate sweep under continuous latent scalar budgets indicates that the visual gains are not explained solely by latent capacity, while also showing that additive metadata embeddings are not a universal source of improvement. Finally, fixed-rate energy selection provides a strong non-parametric baseline: energy_global improves average PSNR over uniform selection by 16.73 dB for audio, 16.90 dB for images, and 15.86 dB for video under compressed keep ratios. Masked sparse training reaches 34.45 dB video PSNR with 50% of dense tokens. The results support a unified wavelet token schema and sparse token interface, while stopping short of establishing a universal discrete vocabulary.
arXiv:2606.03236v1 Announce Type: new
Abstract: Multimodal large language models (MLLMs) have substantially advanced mobile agents, yet proactive mobile assistance remains challenging because agents must decide \emph{when} to intervene before determining \emph{how} to assist. Existing systems often implement these two decisions within a unified MLLM-based pipeline, leading to goal misalignment between conservative intervention filtering and comprehensive assistance generation, as well as redundant inference when the agent should remain silent. To address these limitations, we propose the \textbf{Pre-Reasoning Perception Framework (PRPF)}, a two-stage framework built on perceiving before reasoning. PRPF introduces a lightweight Multimodal Proactive Perceptor (MPP) for intervention gating and context compression, and activates the Proactive Agent Reasoner (PAR) only when intervention is warranted. Experiments on the ProactiveMobile benchmark show that PRPF substantially reduces false trigger rates (FTR) while improving success rates (SR) and inference efficiency over the ProactiveMobile baseline.
arXiv:2606.03918v1 Announce Type: new
Abstract: AI agents can increasingly handle the mechanical tasks of financial analysis: retrieving documents, calculating formulas, updating spreadsheets. The harder, more valuable challenge is reasoning through the open-ended questions that define expert Analyst work. Existing benchmarks do not capture this class of problem, and those that attempt to evaluate open-ended reasoning rely on model-judged outputs that introduce noise and circularity. We present Hedge-Bench 1.0: a benchmark of 102 actual, on-the-job tasks grounded in the explicit reasoning traces of professional hedge fund analysts working with relevant information sources. This approach enables deterministic grading against verified expert steps. Frontier models and agents score below 16\% on the benchmark. We publish the dataset and evaluation harness at github.com/Trata-Inc/trata-hedge-bench.
arXiv:2606.02953v1 Announce Type: new
Abstract: Usage-based theories of grammars posit that creative productivity of the structures of language is both bolstered and constrained by two distinct frequency signals: entrenchment, stemming from high frequency usage, and preemption, stemming from having never observed a particular linguistic structure in a context where one might expect that structure to appear. Large Language Models are also usage-based, in the sense that the structures of language are learned through exposure to vast amounts of text. Here, we test whether or not the opposing statistical forces of entrenchment and preemption also encourage and constrain linguistic productivity in LLMs. We demonstrate across model architectures that larger models recognize and can reproduce with nonce words constructional productivity (entrenchment) in cases of coercion, wherein the broader constructional context coerces an atypical interpretation of a lexical item. However, we also show that even the largest models do not extend negative evidence to novel language, and statistical preemption does not enable models to avoid overgeneralization of patterns that are semantically felicitous, but never observed in data.
arXiv:2603.26025v2 Announce Type: replace-cross
Abstract: Self-organized dynamic patterns in dense active matter are striking manifestations of non-equilibrium physics. A prominent example is the macroscopic elliptical motion observed in quasi-2D bacterial suspensions, which has lacked a physical explanation. Here, we examine a minimal linear response framework coupling bacterial swimming dynamics with fluid flow, treating long-range hydrodynamic interactions as a macroscopic communication channel. We demonstrate that microscopic swim motion, via Jeffery coupling, manifests as a ``phase-leading'' response to local shear flows. System-wide sustained oscillations, on the other hand, require both a critical bacterial density and strict geometric confinement. By analytically predicting the onset cell density and maximum film thickness, our model achieves excellent quantitative agreement with experiments, establishing a unified physical framework for self-organized periodic motion of elongated body in active fluids.
arXiv:2606.02672v1 Announce Type: new
Abstract: Analog In-Memory Computing (AIMC) accelerators execute matrix-vector multiplications directly within memory arrays, reducing data movement and improving DNN inference efficiency. Their limited effective precision motivates heterogeneous architectures that combine analog compute tiles with digital processing units. This letter classifies existing methods for partitioning DNN workloads across these resources by mapping granularity, optimization strategy, and model support, and distills them into a unified four-stage workflow. To demonstrate the workflow on a model class not yet addressed by existing methods, we apply its first two stages to GPT-2, producing the first AIMC-specific precision sensitivity profile for a decoder-only transformer. Sensitivity is dominated by 4 of 49 projections, with the first decoder block's attention output dominating by an order of magnitude. This suggests that projection-level mapping and selective digital execution of early-block and output-facing projections are important for reliable decoder-transformer deployment on AIMC hardware.
arXiv:2501.17377v4 Announce Type: replace
Abstract: Deep Reinforcement Learning (DRL) has emerged as a promising approach for solving Combinatorial Optimization (CO) problems, such as the 3D Bin Packing Problem (3D-BPP), Traveling Salesman Problem (TSP), or Vehicle Routing Problem (VRP), but these neural solvers often exhibit brittleness when facing distribution shifts. To address this issue, we uncover the Satisficing Generalization Edge, which we validate both theoretically and experimentally: identifying a set of promising actions is inherently more generalizable than selecting the single optimal action. To exploit this property, we propose Adaptive Selection After Proposal (ASAP), a generic framework that decomposes the decision-making process into two distinct phases: a proposal policy that acts as a robust filter, and a selection policy as an adaptable decision maker. This architecture enables a highly effective online adaptation strategy where the selection policy can be rapidly fine-tuned on a new distribution. Concretely, we introduce a two-phase training framework enhanced by Model-Agnostic Meta-Learning (MAML) to prime the model for fast adaptation. Extensive experiments on 3D-BPP, TSP, and CVRP demonstrate that ASAP improves the generalization capability of state-of-the-art baselines and achieves superior online adaptation on out-of-distribution instances.
arXiv:2502.02260v2 Announce Type: replace
Abstract: In the past decade, considerable research effort has been devoted to securing machine learning (ML) models that operate in adversarial settings. Yet, progress has been slow even for simple "toy" problems (e.g., robustness to small adversarial perturbations) and is often hindered by non-rigorous evaluations. Today, adversarial ML research has shifted towards studying larger, general-purpose language models. In this position paper, we argue that the situation is now even worse: in the era of LLMs, the field of adversarial ML studies problems that are (1) less clearly defined, (2) harder to solve, and (3) even more challenging to evaluate. As a result, we caution that yet another decade of work on adversarial ML may be failing to produce meaningful progress.
arXiv:2606.03574v1 Announce Type: cross
Abstract: This work proposes a novel solution to predict pulsar timing residuals with limited data, addressing the critical challenge of data scarcity across spin-frequency subgroups of millisecond pulsars in PTA datasets. The proposed solution applies a Long Short-Term Memory (LSTM) network optimized using the model-agnostic meta-learning algorithm, enabling rapid adaptation to new frequency domain by fine-tuning the LSTM network with only a few-shot of ground truth timing residuals. Particle swarm optimization algorithm is also used for automatic hyperparameter optimization, leading to improved prediction accuracy. Our solution, evaluated on the second data release of the International Pulsar Timing Array (IPTA), demonstrates robust generalization with accurate predictions in three metrics across high-frequency test frequency domains, while requiring only 10% of the timing residuals from these domains for model fine-tuning. Furthermore, our lightweight structure only costs 16.86 MB CPU memory and 18 milliseconds for single-step residual prediction. All these characteristics make our solution highly suitable for real-world applications, where effective and real-time predictions of pulsar timing residuals are essential-particularly in resource-constrained environments with limited computational power, memory, or energy availability.
arXiv:2606.02950v1 Announce Type: new
Abstract: The application of AI in education has evolved from monolithic intelligent tutoring systems to a diverse ecosystem of pedagogical agents, including conversational assistants, virtual coaches, and adaptive tutors. This shift requires a unified and scalable data architecture to manage the complex information feedback loops between human instructors, learners, and the varied AI agents. The design, development, and deployment of the data architecture in turn raises a critical issue of validation. This paper addresses this critical need by describing a practical validation strategy for a high-volume data pipeline developed as part of a data architecture for AI-augmented adult learning at the National AI Institute for Adult Learning and Online Education. Our approach involves a two-stage testing methodology to ensure both functional diversity and real-world scalability. First, the QA environment uses a blend of synthetic and real-world data to validate functional correctness across various event types produced from learner and agent interactions. Following this, the production environment successfully processed a total of over 2.7 million production requests across 21 successful runs carrying authentic event data from a large-scale online program. This validation process surfaced crucial insights into data privacy, a key challenge when handling varied data from multiple AI agent data sources. By outlining a replicable testing strategy for a unified data backbone, this research offers a clear framework for institutions and developers aiming to build and support their own heterogeneous suites of AI-powered learning tools. Keywords: Pedagogical Agents, Learning Ecosystems, Data Architecture, Validation, Scalability, Learning Analytics.
arXiv:2606.03121v1 Announce Type: new
Abstract: Multivariate time series forecasting plays a critical role in real-world applications, including weather prediction, stock analysis, and health monitoring. Due to the diversity of data sources, time series exhibit diverse temporal dynamics, often accompanied by various irregularities such as missing values and non-uniform sampling frequencies. Such irregularities lead to complex and asynchronous temporal dependencies across channels. Thus, a single model with a fixed patching scheme often fails to adapt well to diverse multivariate time series, hindering accurate forecasting. In this paper, we propose TiWeaver, a unified framework designed to handle temporal dynamics and fine-grained inter-channel dependencies adaptively. Specifically, we introduce a Graph-Guided Adaptive Tokenizer (G$^2$AT) that divides time series into high contextually coherent patches by jointly considering temporal density and representation consistency. In addition, we propose a Fine-grained Asynchronous Dependency Extractor (FADE), which is designed to model fine-grained asynchronous inter-channel dependencies while incorporating long-term historical dependencies. We evaluate TiWeaver on 12 real-world time series datasets, where it achieves state-of-the-art performance, outperforming existing methods up to 25%. These results demonstrate its robustness and effectiveness across diverse domains and data characteristics.
arXiv:2606.03288v1 Announce Type: new
Abstract: Introductory programming (CS1) courses often struggle to support students' understanding of program execution. While visualizations can make execution processes explicit, their effectiveness depends on design and context, and empirical evidence for AI-generated visualizations remains limited. We propose Generated Animated Traces (GATs), AI-generated, analogy-based, narrated animations that coordinate source code, execution state, and conceptual analogies. We conduct a study at two institutions in CS1 courses (Python, N=961; Java N=151) comparing GATs to textual explanations. We measure immediate learning performance and experience, end-of-course engagement and exam performance. Results show that GATs can yield selective benefits for immediate learning, but benefits are context-dependent and short-term. We observe that GATs' influence on performance is moderated by learner engagement profiles. This finding underscores the importance of personalized approaches.
arXiv:2606.02949v1 Announce Type: new
Abstract: Control barrier functions (CBFs) have become a standard tool in safety critical-control systems. CBFs convert state constraints into real time control conditions that certify forward invariance (meaning that once the system starts in a safe region, it remains there for all future times) and minimally modify a nominal controller only when safety is at risk. In power systems, CBF based methods have been proposed for frequency and voltage safety, but they largely remain disconnected from three key features that are central to power system operation: differential algebraic equation (DAE) models that capture network power flow constraints, safety specifications involving algebraic variables such as bus voltages, and formal verification of the resulting closed loop system. This paper closes this gap by developing a CBF framework for power system DAE models that supports safety constraints on both dynamic and algebraic variables. The framework provides real time safety filtering through an optimization layer that wraps around an existing controller and minimally modifies its command to enforce safety. In addition, it provides formal verification (i.e., a mathematical guarantee that all admissible trajectories satisfy the prescribed safety constraints) through an offline reachability based certificate of safe operation. The result is a unified filter and verify methodology for enforcing and certifying frequency and voltage safety in power systems while preserving the DAE structure of the underlying model.
arXiv:2606.02948v1 Announce Type: new
Abstract: Curvature adaptivity is a classical theme in online optimization: for convex Lipschitz losses, adaptive methods interpolate between the optimal $O(\sqrt{T})$ regret for general convex losses and $O(\log T)$ regret under strong convexity. Recent work has shown that Follow-the-Perturbed-Leader (FTPL) achieves optimal $O(\sqrt{T})$ regret even for online non-convex Lipschitz losses, assuming access to an approximate offline-optimization oracle, but these guarantees do not exploit curvature. We show that FTPL can be made curvature-adaptive in the non-convex setting, without knowing in advance how curvature will accumulate over time. Our algorithm replaces the fixed perturbation scale of standard FTPL with a time-varying scale chosen using only past information. We give a simple follow-the-leader tuning rule for this scale and show that it competes, up to constants, with the best choice in hindsight. The resulting method achieves $O(\sqrt{T})$ regret for arbitrary non-convex Lipschitz losses and improves as cumulative curvature grows; with sufficiently accurate oracle calls, it achieves $O(\log T)$ regret when cumulative curvature grows linearly, which includes the classical strongly convex regime. We complement these upper bounds with matching lower bounds for prescribed cumulative-curvature sequences, already for one-dimensional convex losses, showing that the tradeoff between worst-case non-convex regret and curvature-driven fast rates is intrinsic.
arXiv:2606.03095v1 Announce Type: new
Abstract: AI systems increasingly shape human workflows by generating intermediate artifacts that users can adopt, revise, or ignore. While prior work has shown that AI assistance can improve the efficiency and accuracy of required tasks, less is known about whether it can increase participation in discretionary but beneficial work that users often intend to perform but frequently skip. We study this question in the context of personalized feedback provision in higher education, a pedagogically valuable but often optional practice. We conduct a mixed-methods study combining a randomized field experiment and qualitative interviews in a 300-level machine learning course with n=11 teaching assistants (TAs) and n=88 students. Student submissions were randomly assigned to either (1) a treatment condition where TAs received AI-assisted feedback drafts after grading or (2) a control condition without drafts. TAs remained fully in control and could use, edit, or ignore drafts at their discretion. We find that AI-assisted feedback significantly increases feedback provision (+10.8 percentage points, SE=1.1, p<0.001) and feedback length (+39.8 chars, SE=3.45, p<0.001) without negatively affecting student usefulness ratings or reducing time per character. Qualitative findings suggest that AI-assisted drafts function as editable scaffolds that lower barriers to initiating feedback rather than reducing overall effort. Our findings highlight AI's promise for discretionary but beneficial tasks: increasing work that might otherwise go undone while preserving human control over final outcomes.
arXiv:2606.03059v1 Announce Type: new
Abstract: Microchannel flow boiling is a promising technique for micro-device thermal management, and understanding the bubble dynamics in microchannel flow boiling is important for the applications. Previous studies only focused on single, isolated bubbles, but the bubbles in microchannel flow boiling applications often exist as bubble trains, in which the bubbles interact with each other. Here, we investigate numerically vapor bubble trains in microchannel flow boiling by adopting the flow-focusing technique to form monodispersed bubbles in the upstream of the microchannel. With increasing the initial vapor-liquid volume ratio, the bubble frequency increases while the growth rate of the bubbles decreases because of the reduced bubble size. With increasing the heat flux on the wall or reducing the latent heat of the working fluid, the bubble train growth rate increases because of the increased vaporization rate. The vaporization of the fluid in the upstream causes the bubble expansion and accelerates the bubble movement in the downstream. The wall temperature and the Nusselt number fluctuate because of the periodic pass-through of bubbles.
arXiv:2606.03786v1 Announce Type: new
Abstract: Online debate platforms offer a unique window into the mechanisms driving opinion formation: they capture both explicit political preferences and the peer environment in which those preferences are expressed. In this work, I develop a Bayesian logistic regression model, inspired by ideal point models from political science, to disentangle two competing mechanisms of voting behaviour in online debates: conviction, driven by prior ideological beliefs, and conformity, driven by peer influence. I apply this framework to the Debate.org dataset, comprising approximately 341k votes across 78k debates on 48 socio-political topics. As the debate platform does not provide predefined topic labels for each debate, I infer the topic and stance from the debate text using large language models, and, with a Bayesian approach, I quantify the relative contribution of each mechanism. I find substantial heterogeneity across topics: conviction dominates on issues tied to personal freedoms and lifestyle choices, such as drug legalisation and legalised prostitution, while conformity dominates on several topics widely regarded as paradigmatic cases of moral conviction, including abortion, gun rights, and global warming. These results have implications for the stability of online political discourse and the design of deliberative platforms.
arXiv:2606.02934v1 Announce Type: new
Abstract: Public scientific and metrology releases can leak the hidden settings that produced them. We formalize and quantify this risk as a profiled statistical side-channel audit: a release map exposes finite-band statistics of a power spectral density (PSD), a profiled observer trains labeled template spectra under an explicit budget, and a challenge release is drawn from one of two utility-equivalent recipes separated by a protected coordinate. Averaged PSD bins follow a gamma channel, replaced by a covariance-weighted log-spectrum channel when the bins are correlated; this yields exact Kullback-Leibler divergences, Chernoff exponents, protected-bit advantage bounds, and finite-training, finite-library, finite-compute, and model-mismatch corrections. Our headline result is a finite-band transport-leakage law: after amplitude and blur are eliminated, the protected acid-transport information obeys $I_{\lambda|\alpha,\beta}(K) = (64/1225)\, w \lambda^{6} K^{9} + O(w \lambda^{8} K^{11})$ for $K\lambda \ll 1$, a ninth-order exponent with a closed-form safe band. A step-by-step protocol turns a measured release into these numbers, and a fixed-seed reproducibility package regenerates every table and figure. We instantiate the audit on screened extreme-ultraviolet (EUV) roughness spectra as a model-conditioned case study, with deployment on measured releases the next step.
arXiv:2606.03102v1 Announce Type: new
Abstract: Test-time scaling improves the reasoning performance of large language models but incurs substantial cost in both total computation and latency. Existing adaptive sampling methods partially mitigate this issue by dynamically deciding when to stop sampling, yet they typically rely on heuristic rules or rely on distribution assumptions. In this work, we formulate adaptive sampling as a Markov decision process (MDP). We train a lightweight sampling controller with reinforcement learning (RL) to jointly balance answer correctness, latency, and computation cost. At each round, the controller decides to stop sampling or to acquire additional samples. Our method is lightweight which only relies on statistics of final answers, and can be trained and deployed on CPU. We further show that the resulting framework admits an interpretation as the Lagrangian relaxation of a constrained optimization problem with explicit budget constraints. Experiments against strong baselines such as ASC and ESC show that our method achieves improved trade-offs among answer correctness, sampling rounds, and total samples required.
arXiv:2502.08006v3 Announce Type: replace
Abstract: Training-free guided generation is a widely used and powerful technique that allows the end user to exert further control over the generative process of flow/diffusion models. Generally speaking, two families of techniques have emerged for solving this problem for gradient-based guidance: namely, posterior guidance (i.e., guidance via projecting the current sample to the target distribution via the target prediction model) and end-to-end guidance (i.e., guidance by performing backpropagation throughout the entire ODE solve). In this work, we show that these two seemingly separate families can actually be unified by looking at posterior guidance as a greedy strategy of end-to-end guidance. We explore the theoretical connections between these two families and provide an in-depth theoretical of these two techniques relative to the continuous ideal gradients. Motivated by this analysis we then show a method for interpolating between these two families enabling a trade-off between compute and accuracy of the guidance gradients. We then validate this work on several inverse image problems and property-guided molecular generation.
arXiv:2506.21129v2 Announce Type: replace
Abstract: Autonomous unmanned aerial vehicles (UAVs) increasingly rely on reinforcement learning (RL) for navigation. However, global navigation satellite system (GNSS) spoofing attacks can induce out-of-distribution observation shifts that corrupt value estimation and degrade mission performance. Existing robust RL approaches typically improve resilience against specific attack models but often fail to generalize to attacks not encountered during training. To address this limitation, we propose a curriculum-guided adaptation framework that progressively exposes a robust policy to gradient-based adversarial observation perturbations of increasing intensity while aligning temporal-difference (TD) error distributions across curriculum stages. Rather than adapting to a particular attack model, the proposed approach preserves TD-error consistency to promote transferability across attack conditions. We further derive a TD-space generalization certificate showing that if the TD-error distribution induced by a test-time attack remains sufficiently close to that of the final curriculum stage, the resulting performance degradation is bounded. The framework is evaluated in a UAV deconfliction environment with dynamic 3D obstacles under previously unseen fixed and dynamic GNSS spoofing attacks. Under fixed spoofing conditions, the curriculum-adapted policy achieved near-perfect mission success rates, compared with 20-56% for standard and robust RL baselines. Under dynamic obstacle-luring spoofing attacks, it achieved the highest episodic rewards while reducing mission completion steps by up to 45% across increasing aerial traffic densities.
arXiv:2507.19684v2 Announce Type: replace
Abstract: Socially interactive humanoid robots must engage with humans through their bodies, adapting in real time to a partner's movement, intent, and abilities. This requires models that understand not just how bodies move, but what movement means in a shared social context. Yet evaluation frameworks for interactive motion generation do not measure whether generated follower motion is legible within a shared movement vocabulary, nor whether it is appropriate to the partner's proficiency level. This gap has two causes: existing frameworks rely on kinematic metrics such as FID and beat alignment that cannot measure either property, and existing datasets lack the move annotations and proficiency variation needed. Salsa is well-suited as an evaluation domain: improvised, dyadic, and governed by a move vocabulary and judging criteria covering timing, musicality, technique, difficulty, partnering, and originality. We present CoMPAS3D, a motion capture dataset of improvised partner salsa paired with an evaluation framework covering kinematic quality, two objective metrics (move legibility and proficiency appropriateness), and six competition-based subjective dimensions. The dataset includes 3 hours of improvisation by 18 dancers spanning beginner, intermediate, and professional levels, with over 2,800 expert-annotated segments covering move types, errors, and stylistic elements. We define three benchmarks: move classification (analogous to transcription), proficiency estimation (fluency assessment), and follower generation (dialogue response). Fine-tuned vision-language models perform strongly on objective metrics applied to ground-truth motion sequences. Applied to Duolando and InterGen, the metrics reveal failures that kinematic metrics miss. Human evaluations confirm the gap between generated and ground-truth motion. CoMPAS3D, annotations, benchmark code, and baseline results are publicly available.
arXiv:2502.09755v4 Announce Type: replace
Abstract: Safety-aligned LLMs respond to prompts with either compliance or refusal, each corresponding to distinct directions in the model's activation space. Recent works show that initializing attacks via self-transfer from other prompts significantly enhances their performance. However, the underlying mechanisms of these initializations remain unclear, and attacks utilize arbitrary or hand-picked initializations. This work presents that each gradient-based jailbreak attack and subsequent initialization gradually converge to a single compliance direction that suppresses refusal, thereby enabling an efficient transition from refusal to compliance. Based on this insight, we propose CRI, an initialization framework that aims to project unseen prompts further along compliance directions. We demonstrate our approach on multiple attacks, models, and datasets, achieving an increased attack success rate (ASR) and reduced computational overhead, highlighting the fragility of safety-aligned LLMs. A reference implementation is available at: https://amit1221levi.github.io/CRI-Jailbreak-Init-LLMs-evaluation.
arXiv:2606.03107v1 Announce Type: new
Abstract: This paper develops an impulse-supervised confined exploration framework for learning local optimal controller for a class of nonlinear systems. The proposed approach combines continuous-time approximate dynamic programming (ADP) with an impulsive supervisory layer, where impulsive braking confines the state within a prescribed region in which a local linear approximation of the nonlinear system is valid. This enables desired persistent excitation required for parameter convergence while preventing large state deviations that invalidate local optimality. The resulting hybrid closed-loop system enforces invariance of the exploration region through state-triggered braking inputs. Simulation results on a nonlinear mechanical system demonstrate effectiveness of the proposed approach.