Academic Intelligence · Curated Daily

探索全球前沿学术脉络

AcademicHub 汇聚顶级期刊与预印本平台的实时文献。定制您的专属科研雷达,利用大语言模型自动生成交叉领域文献分析简报。

01.
arXiv (quant-ph) 2026-06-16

Intermodal entanglement in a quantum optical model of HHG due to the back-action on the driving field

arXiv:2603.01315v2 Announce Type: replace Abstract: Preparation of nonclassical light with special quantum properties is essential for quantum technologies. High-harmonic generation (HHG) is a process which not only enables the creation of attosecond pulses but also has the potential to generate light with intricate quantum properties. In a recent experiment [1], nonclassical inter-harmonic correlations have been measured from a HHG source. In this work, we theoretically investigate entanglement between different harmonics within an effective quantum optical model. This model implements a signifcant degree of simplifcation regarding the processes within the target material, treating the material through susceptibilities, as it is usual in quantum optics. Such an approach yields a general description of HHG, permitting the implications that can be derived within it to hold broadly. We find that entanglement is produced as a result of the often neglected back-action. We can qualitatively reproduce experimentally measured nonclassicalities, which suggests that intermodal entanglement can, to an extent, be considered a universal phenomenon associated with HHG, rather than a result of using specific material targets.

02.
arXiv (CS.AI) 2026-06-16

CONCORD: Asynchronous Sparse Aggregation for Device-Cloud RAG under Document Isolation

arXiv:2606.15179v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) has emerged as a pivotal technique for improving language models by incorporating external knowledge at inference time. As device-cloud collaborative inference makes it feasible to deploy small language models on edge devices, a new setting arises in which private documents remain on the device and public knowledge resides in the cloud. Privacy and policy constraints often forbid raw document exchange, creating a document-isolated dual-end RAG setting. However, existing methods rely on frequent remote synchronization and dense evidence transfer, limiting throughput under realistic latency and bandwidth conditions. To address this issue, we propose CONCORD, an asynchronous sparse aggregation framework for dual-end RAG under document isolation. CONCORD treats the cloud as an asynchronously arriving evidence source rather than a continuously synchronized co-generator. Specifically, we introduce waiting debt control to decide whether each decoding step should continue waiting for remote participation based on the observed return of waiting. We also design a certificate-guided minimal supplementation mechanism that requests only the remote evidence needed to determine the current greedy decision. Steps that consult the cloud preserve the same greedy token as dense dual-end aggregation, while the remaining steps commit locally without remote evidence. Experiments on Natural Questions and WikiText-2 show that CONCORD improves end-to-end throughput over baselines by $1.66\times$ and $2.15\times$, respectively, while reducing per-token communication by over two orders of magnitude and maintaining comparable answer quality and perplexity.

03.
arXiv (CS.AI) 2026-06-12

From AGI to ASI

arXiv:2606.12683v1 Announce Type: new Abstract: Over the last decade, building human-level artificial general intelligence has moved from far-fetched speculation to being a concrete next-decade target for many of the largest AI organisations. Achieving this goal would have profound and far-reaching impacts on human society, which raises many complex questions for the decade ahead. This report investigates how AI itself might continue to develop in a post-AGI world along the continuum of machine intelligence. The endpoint of this continuum, Universal AI, is theoretically well understood, which provides some formal grounding for the main focus of this report: the transition from human-level AGI to artificial general superintelligence, which, intuitively, can be understood as a system that is more intelligent and cognitively capable than large organisations of humans. After characterizing ASI, the report discusses four potential pathways from AGI to ASI: scaling AGI, AI paradigm shifts, recursive improvement, and ASI emerging from large-scale multi-agent collectives. The report then discusses possible frictions and bottlenecks along these pathways. Determining whether the impact of these frictions will be negligible or substantial raises a number of concrete open research questions. Due to large uncertainties for predicting ASI progress, it cannot be ruled out that AI progress might continue to accelerate over the next years. This could imply that the image of a single transformative step change, caused by the introduction of human-level AGI into our society, could be inaccurate. More apt might be the prospect of a series of transformative societal changes caused by AI-enabled progress and breakthroughs across many areas of science and technology. Preparing for this prospect requires a massively interdisciplinary endeavour of global scope and interest.

04.
arXiv (CS.AI) 2026-06-19

The Autonomy Tax: Defense Training Breaks LLM Agents

arXiv:2603.19423v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents increasingly rely on external tools (file operations, API calls, database transactions) to autonomously complete complex multi-step tasks. Practitioners deploy defense-trained models to protect against prompt injection attacks that manipulate agent behavior through malicious observations or retrieved content. We reveal a fundamental capability-alignment paradox: defense training designed to improve safety systematically destroys agent competence while failing to prevent sophisticated attacks. Evaluating defended models against undefended baselines across 97 agent tasks and 1,000 adversarial prompts, we uncover three systematic biases unique to multi-step agents. Agent incompetence bias manifests as immediate tool execution breakdown, with models refusing or generating invalid actions on benign tasks before observing any external content. Cascade amplification bias causes early failures to propagate through retry loops, pushing defended models to timeout on 99\% of tasks compared to 13\% for baselines. Trigger bias leads to paradoxical security degradation where defended models perform worse than undefended baselines while straightforward attacks bypass defenses at high rates. Root cause analysis reveals these biases stem from shortcut learning: models overfit to surface attack patterns rather than semantic threat understanding, evidenced by extreme variance in defense effectiveness across attack categories. Our findings demonstrate that current defense paradigms optimize for single-turn refusal benchmarks while rendering multi-step agents fundamentally unreliable, necessitating new approaches that preserve tool execution competence under adversarial conditions.

05.
bioRxiv (Bioinfo) 2026-06-12

CAREPath: Semantic Context-Aware Reasoning Paths with Mechanism-Augmented Embeddings for Drug Repurposing

Biomedical knowledge graphs (BKGs) that include drugs, genes, and diseases support drug repurposing by connecting drugs to diseases through gene-mediated multi-hop paths, thereby enabling mechanism-of-action reasoning. However, deeper traversal does not necessarily improve mechanistic reasoning: long paths grow combinatorially and frequently pass through hub genes, producing irrelevant gene regulatory signals, whereas overly constrained or sparse paths may miss broader biological context. We propose CAREPath, a KG-LLM framework inspired by depth-first search (DFS)-like and breadth-first search (BFS)-like reasoning to balance mechanistic specificity, scalability, and context recovery. The DFS-like module constrains traversal to short disease-gene-drug paths, converts each path into a structured prompt, and encodes it with a biomedical language model to generate semantic path embeddings. Complementarily, the BFS-like module constructs entity-level mechanism-context embeddings from one-hop gene neighborhoods and enriches them through similarity-guided augmentation using pharmacologically related drugs and gene-signature-similar diseases. Across five biomedical KGs, CAREPath achieves the best overall AUPRC among 18 baselines, improving performance by up to 3.8%. Additional analyses show that semantic short-path encoding contributes most to performance, while mechanism-context augmentation improves robustness under sparse evidence and strengthens Gene Ontology functional agreement. Case studies and recently FDAapproved indications further demonstrate its practical relevance, positioning CAREPath as an interpretable framework for scalable and mechanism-aware drug repurposing. Source code is available at https://github.com/hamppy-song/CAREPath.

06.
arXiv (CS.AI) 2026-06-16

Localizing Credit at the Divergence: Path-Conditioned Self-Distillation for LLM Reasoning

arXiv:2606.15576v1 Announce Type: cross Abstract: Reinforcement learning from verifiable rewards assigns a single scalar to each rollout, leaving token-level credit assignment underspecified in long reasoning traces. On-policy self-distillation addresses this by letting the same model act as a teacher conditioned on privileged information, producing a dense per-token signal. But the common choice of a ground-truth answer is only an endpoint cue: on terse-answer tasks, the teacher falls silent at the intermediate positions where path-level guidance matters most. We propose Hindsight Self-Distillation (HSD), which conditions the teacher on a successful peer rollout drawn from the current training group. Such a peer is an exact sample from the success-conditioned policy, requiring no additional sampled rollouts. By providing a full successful continuation rather than only the final answer, the resulting credit signal concentrates at the divergence position between a failed rollout and a successful peer. Across Qwen3-8B and Qwen3-32B on math and code benchmarks, HSD obtains the best result against GRPO variants and on-policy distillation baselines, with the largest gains on terse-answer tasks such as AIME.

07.
arXiv (CS.LG) 2026-06-19

Linear Mode Connectivity under Data Shifts for Deep Ensembles of Image Classifiers

arXiv:2511.04514v2 Announce Type: replace Abstract: The phenomenon of linear mode connectivity (LMC) links several aspects of deep learning, including training stability under noisy stochastic gradients, the smoothness and generalization of local minima (basins), the similarity and functional diversity of sampled models, and architectural effects on data processing. In this work, we experimentally study LMC under data shifts and identify conditions that mitigate their impact. We interpret data shifts as an additional source of stochastic gradient noise, which can be reduced through small learning rates and large batch sizes. These parameters influence whether models converge to the same local minimum or to regions of the loss landscape with varying smoothness and generalization. Although models sampled via LMC tend to make similar errors more frequently than those converging to different basins, the benefit of LMC lies in balancing training efficiency against the gains achieved from larger, more diverse ensembles. Code and supplementary materials are available at https://github.com/DLR-KI/LMC. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.

08.
arXiv (CS.CV) 2026-06-11

4DP-QA: Scalable QA for 4D Perception in Vision Language Models

Despite recent advances, Vision Language Models (VLMs) still struggle to grasp the dynamics of the world. We note that the ability to reason about a 4D scene, challenging in itself, is further complicated by two factors. First, VLMs observe motion indirectly via its projection onto 2D images. Second, existing datasets fail to disentangle object and camera motion. To address these challenges, we present a QA generation pipeline that focuses on motion-related scene understanding. We take particular care of the entanglement of camera and object motion by casting tracking in both the traditional way and in a novel, fixed reference system, dubbed True-Motion Tracking, which provides an intuitive description of motion. From this pipeline, we generate a large-scale training dataset of 400K samples, 4DP-QA (4D Perception QA), and a 2.2K-sample benchmark, 4DP-QA-Bench. Training existing models on our dataset yields performance improvements on an external benchmark, validating the effectiveness of our method.

09.
arXiv (CS.LG) 2026-06-18

FinP: Fairness-in-Privacy in Federated Learning by Addressing Disparities in Privacy Risk

arXiv:2502.17748v4 Announce Type: replace Abstract: Federated Learning (FL) inherently mitigates mass data centralization risks; however, its privacy protections are not equally distributed - leaving vulnerable individuals disproportionately exposed to sophisticated privacy attacks. Crucially, statistical heterogeneity in human-centric FL environments often results in an inequitable distribution of privacy risks, particularly affecting those whose sensitive attributes or behaviors make them outliers. To address this critical gap, we introduce FinP, a novel framework designed to formalize and enforce fairness-in-privacy by mitigating disproportionate client vulnerability to Source Inference Attacks (SIA). FinP operationalizes a two-pronged defense strategy that tackles both the symptoms and root causes of privacy disparity, ensuring that no group of clients bears an excessive privacy burden. It combines a server-side adaptive aggregation mechanism, which dynamically weights client contributions based on their estimated privacy risk, with a client-side regularization technique to curb localized overfitting that drives unique data memorization. Extensive empirical evaluations on FEMNIST, Human Activity Recognition (HAR), and CIFAR-10 datasets demonstrate that FinP effectively aligns privacy fairness with primary task utility. Notably, FinP successfully mitigates SIA risks and reduces disparities in privacy exposure, establishing that strong fairness-in-privacy guarantees need not compromise model utility. Ultimately, FinP establishes equitable privacy protections by reducing vulnerability disparities by up to 57.14%, while preserving global model utility within a marginal +/- 1.75% of standard federated baselines.

10.
arXiv (CS.CL) 2026-06-15

An Empirical Study of Automating Agent Evaluation

Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive. A natural question arises: can frontier coding assistants reliably automate this evaluation process? Our study shows that simply prompting coding assistants is insufficient for this task. Without domain-specific evaluation knowledge, frontier coding assistants achieve only a 30% execution success rate and produce over-engineered evaluations averaging 12+ metrics per agent, indicating that strong coding ability does not automatically translate to reliable agent evaluation. We introduce EvalAgent, an AI assistant that automates the end-to-end agent evaluation pipeline. EvalAgent encodes evaluation domain expertise as evaluation skills (procedural instructions, reusable code and templates, and dynamically retrieved API documentation) that compose into a trace-based pipeline producing complete evaluation artifacts including metrics, executable code, and reports. To systematically assess generated evaluations, we introduce a meta-evaluation framework alongside AgentEvalBench, a benchmark comprising 20 agents, each paired with evaluation requirements and test scenarios. We further propose the Eval@1 metric to measure whether generated evaluation code both executes and yields meaningful results on the first run. Our experiments show that EvalAgent produces focused evaluations, improving Eval@1 from 17.5% to 65%, and achieving 79.5% human expert preference over baseline approaches. Further ablation studies show that evaluation skills are critical for handling complex evaluation: removing them causes Eval@1 to drop significantly from 65% to 30%.

11.
arXiv (CS.LG) 2026-06-12

Adaptive Model-Predictive Control of a Soft Continuum Robot Using a Physics-Informed Neural Network Based on Cosserat Rod Theory

arXiv:2508.12681v3 Announce Type: replace-cross Abstract: Dynamic control of soft continuum robots (SCRs) holds great potential for expanding their applications, but remains a challenging problem due to the high computational demands of accurate dynamic models. While data-driven approaches like Koopman-operator-based methods have been proposed, they typically lack adaptability and cannot reconstruct the full robot shape, limiting their applicability. This work introduces a real-time-capable nonlinear model-predictive control (MPC) framework for SCRs based on a domain-decoupled physics-informed neural network (DD-PINN) with adaptable bending stiffness. The DD-PINN serves as a surrogate for the dynamic Cosserat rod model with a speed-up factor of up to 44,000. It is also used within an unscented Kalman filter for estimating the model states and bending compliance from end-effector position measurements. We implement a nonlinear evolutionary MPC running at 70 Hz on the GPU. In simulation, it demonstrates accurate tracking of dynamic trajectories and setpoint control with end-effector position errors below 3 mm (2.3\% of the actuator's length). In real-world experiments, the controller achieves similar accuracy and accelerations up to 3.55 m/s2.

12.
arXiv (CS.CV) 2026-06-16

RQUL-UIE: Revitalizing Quality-Unstable Labels for Underwater Image Enhancement via In-Dataset Self-Supervision

Underwater Image Enhancement (UIE) is essential for mitigating degradations caused by water medium. Although learning-based methods have advanced significantly, most rely on paired datasets with unstable label quality, which bottlenecks model performance. This paper proposes a diffusion-based, in-dataset self-supervised learning strategy designed to exploit the quality distribution of training labels. Specifically, we evaluate label quality via semantic perception embeddings from a pre-trained diffusion model in a training-free manner. These quality scores are subsequently quantized into noise-level indices, guiding a multi-step denoising process for level-wise supervision. This mechanism prevents low-quality labels from degrading the model while maximizing their utility during training. Furthermore, a Fourier-based refinement network is incorporated to explicitly reconstruct high-frequency components. Extensive evaluations demonstrate that our method consistently outperforms SOTA approaches in restoration quality. The code and pre-trained model will be available once accepted in link.

13.
arXiv (CS.AI) 2026-06-16

Demystifying Variance in Circuit Discovery of LLMs

arXiv:2606.16920v1 Announce Type: cross Abstract: Circuit discovery is a key technique in mechanistic interpretability to pinpoint the model components that are crucial for performing a given task. Although the current state-of-the-art method (EAP-IG) performs well on the metric of (un)faithfulness, it suffers from substantial variability. This includes resampling variance, where the circuit changes when we probe with a new batch of data from the same distribution; rephrasing variance, where the discovered circuit shifts when the prompts are rephrased; and sample-wise variance, where a circuit with low population unfaithfulness exhibits large fluctuations in unfaithfulness across individual samples. This paper studies the roots of these variances. We demonstrate that CEAP, our new circuit discovery method that improves upon EAP-IG with a theoretical guarantee, can substantially lessen resampling variance. We further show that rephrasing variance arises because prompts with different templates tend to activate different circuits in the model. This leads us to argue that it may be challenging to find a comprehensive circuit that explains and controls the model's behavior on a task, which can be expressed in countless templates, suggesting that LLMs may be inherently hard to steer. We show that sparsity, which has been claimed to form more compact and interpretable task circuits, fails to solve this problem. Regarding sample-wise variance, we argue that it is largely benign: extremely poor unfaithfulness scores often stem from how unfaithfulness is defined, rather than from defects in the measured circuits. We show that the magnitude of unfaithfulness is affected by selective contribution scaling, a neural mechanism that accounts for the extremely poor scores sometimes observed.

14.
arXiv (CS.CL) 2026-06-16

ACCORD: Action-Conditioned Contextual Grounding for Language Agents

User instructions are often underspecified because humans rely on implicit assumptions about the surrounding environment. For large language model (LLM) agents operating in information-rich digital and physical environments, these assumptions cannot be inferred from the instruction alone; they must be recovered from the current state of tools, data, interfaces, and observations. Effective execution therefore requires agents to identify missing context, ground it in observed evidence, and carry it forward into subsequent actions. We show that current agents often fail to do so. They act from assumed rather than observed specifics, overlook information they could have gathered, and fail to incorporate evidence that has already been returned. Building on this insight, we propose ACCORD (Action-Conditioned Contextual Grounding), a simple and effective agent framework for adaptive grounding. Before each action, ACCORD actively probes the environment for missing information and integrates relevant context from the agent's trajectory that would otherwise be overlooked. Requiring no additional training or task-success signals, ACCORD improves task-goal completion on AppWorld by up to +20.6 points with GPT-5-mini, from 42.0% to 62.6%, compared to strong baselines. These gains persist with a substantially stronger base model (+10.8 with Claude-4.5-sonnet), an open-weight model (+10.1 with Qwen3.5-27B-FP8), and on the embodied AlfWorld benchmark (+7.4 success rate with GPT-5-mini).

15.
arXiv (CS.AI) 2026-06-15

The Weight Norm Sets the Grokking Timescale: A Causal Delay Law

arXiv:2606.13753v1 Announce Type: cross Abstract: Grokking is the delayed onset of generalization in neural networks, arising long after they fit the training data. Whether the weight norm causes this delay is disputed: some studies report a critical norm at the transition, others observe grokking with no fixed norm at all. We settle this by intervening on the norm during training rather than only observing it. Under free training with weight decay, networks grok when the weight norm reaches a value Wc that varies little across seeds and learning rates (CV 1 to 2 percent) and grows with the modular base as a power law. When we instead clamp the norm to a fixed multiple rho of Wc and hold it there, the network still groks, but the delay follows T_grok proportional to exp(alpha rho). One exponent, alpha near 7.5, fits this delay across four moduli (R^2 = 0.996). Over the swept ranges the held norm moves the delay by about 19x and the learning rate by only about 2x, and holding the norm above Wc slows grokking rather than preventing it. A final LayerNorm removes the dependence by decoupling weight scale from the network function; without it the exponential law returns. This pinned-norm delay is the exponential counterpart to the logarithmic delay predicted for a freely contracting norm.

16.
arXiv (CS.AI) 2026-06-15

FPGA-Based Neural Network Accelerators for Space Applications: A Survey

arXiv:2504.16173v3 Announce Type: replace-cross Abstract: Space missions are becoming increasingly ambitious, necessitating high-performance onboard spacecraft computing systems. In response, field-programmable gate arrays (FPGAs) have garnered significant interest due to their flexibility, cost-effectiveness, and radiation tolerance potential. Concurrently, neural networks (NNs) are being recognized for their capability to execute space mission tasks such as autonomous operations, sensor data analysis, and data compression. This survey serves as a valuable resource for researchers aiming to implement FPGA-based NN accelerators in space applications. By analyzing existing literature, identifying trends and gaps, and proposing future research directions, this work highlights the potential of these accelerators to enhance onboard computing systems.

17.
arXiv (CS.CV) 2026-06-15

Self-Evolving Visual Questioner

Vision-language models (VLMs) are typically trained as passive answerers, while their ability to actively ask diverse, non-trivial, visual-centric and grounded questions remains underexplored. Existing visual questioners' performance is bottlenecked by the availability of high-quality training data or the cost of curating them. We show that a VLM can continuously improve itself as a visual questioner without any external supervision. We propose a self-evolving framework that uses a VLM itself as both a proposer and a filter to produce harder, more informative, and visual-centric questions, while maintaining their exploration diversity to avoid training collapse. These questions are then used to train the VLM in both questioner and answerer modes. To evaluate the questioner, we introduce an agentic protocol that assesses questions along perception, reasoning, and diversity dimensions. Experiments across various backbone VLMs show that our method substantially enhances the quality and substantially expands the difficulty boundary of autonomous question generation. Under the same budget, our self-supervision is more effective than training on the static source data. Moreover, the self-evolving questioner remains a competitive or even better answerer.

18.
arXiv (CS.CL) 2026-06-18

STARE: Surprisal-Guided Token-Level Advantage Reweighting for Policy Entropy Stability

Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training. We conduct a first-order gradient analysis of token-level entropy dynamics under GRPO and identify a token-level credit assignment mismatch: the per-token entropy variation decomposes into the product of the trajectory-level advantage and an entropy sensitivity function over the next-token distribution, yielding an advantage-surprisal four-quadrant structure and a near-criticality property. Motivated by it, we propose STARE (Surprisal-guided Token-level Advantage Reweighting for policy Entropy stability), which identifies entropy-critical token subsets via batch-internal surprisal quantiles, selectively reweights their effective advantages, and incorporates a target-entropy closed-loop gate for stable entropy regulation. Across model scales from 1.5B to 32B and three task families (Short CoT, Long CoT, and Multi-Turn Tool Use), STARE sustains stable RL training over thousands of steps while maintaining policy entropy within the target band. On AIME24 and AIME25, STARE outperforms DAPO and other competitive baselines by 4%-8% in average accuracy, with reflection tokens and response length growing in tandem, indicating sustained exploration-exploitation balance that further unlocks RL training potential.Code is available at https://github.com/hp-luo/STARE.

19.
arXiv (CS.CV) 2026-06-16

Disagreement-Based Cross-Model Routing for Implicit Video Question Answering

We study multiple-choice video question answering on the ImplicitQA benchmark, where the correct answer is never explicitly shown but must be inferred from off-screen events, line-of-sight cues, causal structure, and cross-shot spatial layout. On this benchmark a single frontier video LLM already operates near its accuracy ceiling, and we observe that conventional self-consistency strategies – majority voting across repeated samples of the same model – can hurt rather than help, because the model's errors on hard questions are correlated. We propose disagreement-based cross-model routing, a pure inference-time procedure that requires no labels and no training. We triple-sample a native-video model (Gemini 3.1 Pro Preview) at temperature zero, exploit the genuine sample-to-sample variance of its video-processing pipeline to identify the roughly 20% subset of questions where the three samples disagree, and route only that subset to a second model from a different family (Claude Opus 4.8) that consumes uniformly sampled frames with adaptive thinking. On the 1001-question validation set with public ground truth – our main evaluation – the method improves AvgAcc by +1.43 over the best single sample of the primary model, with per-category gains concentrated on Motion & Trajectory (+5.49), Inferred Counting (+3.45), and Vertical Spatial Reasoning (+1.82) – the categories most dependent on cross-shot reference resolution. The same pipeline applied to the held-out 172-question CVPR 2026 ImplicitQA challenge test set achieves 82.03 AvgAcc / 79.71 MacroAvgAcc (+1.81 over the best single sample of the primary model), confirming the validation result on an independent split.

20.
arXiv (CS.LG) 2026-06-19

FloatDoor: Platform-Triggered Backdoors in LLMs

arXiv:2606.19535v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed in sensitive settings such as software engineering, where their outputs directly shape downstream artifacts. Recent work has shown that an identical model can produce measurably different outputs depending on the deployment platform, a consequence of non-associative floating-point arithmetic and divergent kernel implementations. We study the security implications of this platform-dependent variability and uncover a novel attack surface on LLM deployments. We introduce FloatDoor, the first input-independent, platform-triggered backdoor attack against generative LLMs. The compromised model exhibits adversary-chosen behavior when served on a target platform and is otherwise benign. FloatDoor is realized through two lightweight LoRA adapters, one that amplifies inter-platform numerical divergence and one that binds the resulting platform signature to a malicious downstream task, while leaving aggregate model utility largely intact. FloatDoor exploits a pronounced time-of-check, time-of-use gap between model auditing and serving. We demonstrate FloatDoor on Qwen3-4B across a broad range of deployment targets, including NVIDIA GPUs, Google TPUs, AWS Graviton, and Alibaba Yitian-710. As a final case study, we show that FloatDoor reliably induces exploitable code vulnerabilities on a chosen target platform. Our results establish a new class of attacks on LLM deployments and underscore the pressing need for trusted model supply chains in sensitive, LLM-powered applications.

21.
arXiv (CS.AI) 2026-06-12

Creating and Evaluating K-12 GenAI Assessment Graders Through Context Engineering

arXiv:2606.12422v1 Announce Type: cross Abstract: The integration of large language models (LLMs) into educational assessment represents a transformative shift in classroom grading practices. While automated scoring systems and machine learning techniques have existed for decades, generative AI (GenAI) now enables educators to implement standards-based grading (SBG) with unprecedented efficiency and scale. This paper examines the theoretical foundations and evaluates an LLM grader that uses commercially available foundation models with context and prompt engineering to score student work against a rubric. Drawing on an empirical interrater agreement study using Massachusetts Comprehensive Assessment System (MCAS) data, we observed the Quadratic Weighted Kappa (QWK) and Proportional Reduction in Mean-Squared Error (PRMSE) across mathematics, science, and ELA, using Claude Sonnet 4, Haiku 4.5, GPT-5, and GPT-5 Mini. The results demonstrate that LLM graders, especially when based on foundational models with more parameters, achieve substantial agreement with human raters in mathematics and science assessments, while the performances vary in ELA, suggesting generic foundation models can be effective at scoring in given contexts. Additional analysis of teacher and student feedback reveals strong acceptance of AI-generated narrative feedback but skepticism toward numerical scores, suggesting that LLMs function most effectively as formative tools rather than summative evaluators. Our findings indicate that thoughtfully designed hybrid models that combine AI efficiency with teacher judgment can reduce workload, enhance feedback quality, and support equitable assessment practices without displacing professional expertise.

22.
arXiv (CS.AI) 2026-06-11

When Do Data-Driven Systems Exhibit the Capability to Infer?

arXiv:2606.11769v1 Announce Type: new Abstract: The European AI Act is the first comprehensive regulation of artificial intelligence (AI), setting out extensive obligations, particularly for so-called high-risk and general-purpose AI systems. A key distinguishing feature of AI systems under the AI Act is the capability to infer. Since the AI Act does not clearly define what inference is, there is a gray area for certain data-driven systems. A specific example is credit scoring systems, which are listed by Annex III of the AI Act. At the same time, however, these are often implemented using statistical models for which it is unclear whether they have the capability to infer and thus fall under the AI definition of the AI Act at all. Motivated by statistical learning theory, this work develops a framework for grading different levels of the capability to infer. Based on the AI Act and the Commission Guidelines on the definition of an artificial intelligence system, we analyze which levels constitute sufficient capability to infer within the meaning of the AI Act and where further regulatory clarity is needed. We illustrate the framework by creating two realistic credit scoring workflows and show whether and where inference occurs in them. Our analysis illustrates that not only individual models but the entire data processing workflow must be considered. It also shows that the involvement of human experts during development can have significant influence on the capability to infer. Code can be found at https://github.com/fraunhofer-iais/inference-framework-creditscorecards.

23.
arXiv (CS.CL) 2026-06-17

Learning from the Self-future: On-policy Self-distillation for dLLMs

On-policy self-distillation (OPSD) has proven effective for post-training large language models (LLMs), yet its application to diffusion LLMs (dLLMs) remains unexplored. Existing OPSD methods are inherently autoregressive-centric. They inject privileged information via left-to-right prefix conditioning with token-level divergence supervision, a design that fundamentally conflicts with the arbitraryorder generation of dLLMs. We introduce d-OPSD, the first OPSD framework tailored for dLLMs. Our approach makes two core contributions. First, we reframe self-teacher construction by using self-generated answers as suffix conditioning, enabling the student model to learn from "self future-experience" rather than privileged prefixes. Second, we shift supervision from token-level to step-level, aligning training with the iterative denoising process of dLLMs. Experiments across four reasoning benchmarks show that d-OPSD consistently outperforms RLVR and SFT baselines with superior sample efficiency, requiring only around 10% of the optimization steps by RLVR and opening a promising pathway for dLLM posttraining. The code is available at https://github.com/xingzhejun/d-OPSD.

24.
arXiv (quant-ph) 2026-06-17

Efficient time-series prediction on NISQ devices via time-delayed quantum extreme learning machine

arXiv:2602.21544v2 Announce Type: replace Abstract: We proposed a time-delayed quantum extreme learning machine (TD-QELM) for efficient time-series prediction on noisy intermediate-scale quantum (NISQ) devices. By encoding multiple past inputs simultaneously, TD-QELM achieves shallow circuit depth independent of sequence length, thereby, mitigating noise accumulation and reducing computational complexity. Experiments using the NARMA benchmark on both noiseless simulations and IBM's 127-qubit processor demonstrate that TD-QELM consistently outperforms conventional quantum reservoir computing in prediction accuracy and noise robustness. These results highlight TD-QELM as a practical and scalable framework for time-series learning on current NISQ hardware.

25.
arXiv (CS.CL) 2026-06-16

Evaluating the Robustness of Proof Autoformalization in Lean 4

Proof autoformalization aims to translate a mathematical informal proof written in natural language into a formal proof in a formal language such as Lean~4. Several works have developed LLM-based models for proof autoformalization. However, existing evaluations have typically focused on translating well-formed informal proofs from curated datasets. We argue that a robust proof autoformalizer must remain faithful even for informal proofs that diverge from these idealized ones, and we present the first study on the robustness of proof autoformalization models. We formulate two categories of perturbations and evaluate robustness under each: a global perturbation paraphrases the informal proof in a different style, under which the formalization should remain consistent; a local perturbation alters a value, symbol, or proof step, possibly in a counterfactual way, and a robust formalization should faithfully reflect the perturbation rather than reverting to the original one or inferring a different one on its own. We build a benchmark with both perturbations on miniF2F and MATH-500, and automatically measure how stable a proof autoformalization's correctness is under global perturbations and how faithfully its output reflects local perturbations. We evaluate seven recent models, all of which are sensitive to global perturbations and mostly fail to remain faithful under local perturbations. Code and data are available via https://github.com/ucr-rai/robust-proof-autoformalization.