Papers

Industrial monitoring models must detect operationally relevant deviations while satisfying target-specific data, calibration, and resource constraints. Time-series foundation models (TSFMs) promise reusable representations and zero-shot forecasts, yet evidence for their deployment value remains mixed when task definitions are heterogeneous and lightweight baselines are competitive. This work presents a protocol-aware empirical assessment across three settings: a C-MAPSS degradation-risk proxy, normal-only training for anomalous-sound detection on MIMII, and BDG2 forecasting-residual diagnostics with synthetic target perturbations. We assess classical one-class methods, compact neural autoencoders, residual forecasters, MOMENT-small, Chronos-T5, and TimesFM 2.5 in terms of anomaly-ranking performance, risk-horizon sensitivity, residual forecasting and perturbation sensitivity, and local implementation cost. Across 100 C-MAPSS engines evaluated out of fold, TCN-AE reaches fold-weighted AUROC/AUPRC 0.9570/0.8960, compared with 0.7310/0.3080 for MOMENT reconstruction; paired engine-cluster bootstrap confidence intervals exclude zero for both differences. Across five matched MIMII pump evaluations, OCSVM also exceeds MOMENT reconstruction in AUROC and AUPRC. On a fixed 12-meter BDG2 panel, TimesFM 2.5 has the lowest aligned forecast error and the highest synthetic AUROC point estimate, although synthetic AUPRC is similar across TSFM and fitted residual models. Same-device measurements show that MOMENT incurs higher latency, peak allocated VRAM, and serialized state-dictionary size than TCN-AE. Under the evaluated frozen and zero-shot settings, TSFMs are task-dependent deployment options rather than default replacements for fitted lightweight models.

Practical molecular inverse design is rarely a one-shot generation problem; it often involves enriching a pool of molecules that match a desired property profile under a limited oracle budget. Bayesian optimization (BO) offers a natural framework for this setting, with standard Gaussian-process (GP) surrogates operating in continuous embedding spaces. We propose LoopMol, a closed-loop framework whose large language model (LLM) surrogate reasons directly over task instructions, SMILES-level optimization histories, and oracle feedback. At each iteration, the surrogate selects informative reference molecules under an exploration–exploitation principle, optionally with a concise guidance sentence. These outputs condition the next generation round and yield an inspectable optimization trace. Experiments on six drug and material tasks across three Llamole backbones show that the LLM-based variant improves over one-shot generation and is competitive with or stronger than GP-based BO. They also reveal a domain-dependent interface: reference-only conditioning performs better on binary drug tasks, while adding a concise guidance sentence generally benefits continuous material tasks.

Accurate extrinsic calibration between event-based and frame-based cameras remains a practical bottleneck for heterogeneous stereo systems. Existing approaches often require sensor or target motion, precise synchronization, or computationally expensive event-to-image reconstruction. We propose a simple, motion-free cross-modal calibration framework that uses a temporally modulated, blended ChArUco target presented on standard consumer displays. By alternating between the original pattern and a partially blended version, the target reliably triggers events while remaining continuously observable to a frame-based camera, avoiding blank frames and reducing synchronization constraints to a coarse, trigger-based alignment. We discretize events into frames coarsely aligned with the RGB images, apply lightweight denoising, and perform ChArUco-based intrinsic and stereo extrinsic calibration. Extensive experiments assess robustness to blending opacity, display brightness, external illumination, viewing angle, and handheld acquisition. Compared to the strongest motion-based reference (E2Calib + Kalibr) and a non-motion-based reference (Plasberg et al.), our approach reduces the mean reprojection error by $44\%$ and $6\%$, respectively, while substantially simplifying the calibration procedure. Finally, we demonstrate practical utility in a robotic eye-to-hand calibration case study, showing consistent transformations and stable downstream geometric measurements even under partial occlusions. Code is publicly available at https://github.com/nhessenthaler/simple-evrgb-cal.

The growing availability of commodity RISC-V hardware has sparked interest in its use for High Performance Computing (HPC), with PCIe accelerator cards offering a practical near-term pathway to adoption. The Tenstorrent Wormhole is one example, with dedicated vector and matrix units across 128 Tensix cores, and is widely available. In this paper, we explore porting the AX kernel of Nekbone, a widely used HPC mini-application derived from the Gordon Bell Prize-winning Nek5000 spectral element solver, onto the Wormhole accelerator. This kernel evaluates the Poisson operator, and we describe the mapping of the algorithm onto the Tensix. The initial performance results reveal that the host-side data transposition, required for the z-direction gradient computation, is a severe bottleneck. Consequently, we investigated two optimisation strategies that yield dramatic improvements, achieving 242.97 GFLOPS for 100000 elements across 128 Tensix cores, outperforming a 24-core Xeon Platinum CPU and drawing approximately 7 times less power.

Multimodal Large Language Models (MLLMs) are increasingly deployed as multi-step agents, where explicit reasoning supports task decomposition and tool coordination but also accumulates self-generated text. Over long trajectories, this text can dominate the context and suppress visual evidence, creating textual debt. We observe that reasoning becomes redundant once task-relevant visual evidence is grounded, while stale hypotheses can misguide later inference when grounding remains uncertain. Pruning must therefore remove redundant text without discarding visual evidence. We propose SPARE, a Kullback-Leibler (KL)-guided framework for pruning accumulated reasoning in multimodal tool-use agents. SPARE uses a compact task-state summary as privileged diagnostic context. For each candidate segment, it replays the same model under the original and summary-conditioned contexts. Reverse-KL divergence from on-policy self-distillation (OPSD) then tests whether the summary sufficiently covers the segment without disrupting future reasoning. We further fine-tune the summarizer with supervised fine-tuning (SFT), enabling more compact summaries, broader coverage, and more aggressive pruning. Across multi-step visual tool-use benchmarks, SPARE achieves the highest average accuracy among pruning methods while removing 37.89-64.58% of reasoning tokens. This favorable accuracy-context trade-off shows that reducing textual dominance restores reliance on visual evidence and mitigates over-conditioning on self-generated language.

Coding agents are increasingly evaluated not only by whether they solve a task, but also by how they execute it. However, existing process-level evaluations often treat action prediction, task uncertainty, and step attribution as if they were the same problem, which makes it unclear what such evaluations actually measure. In this paper, we introduce a measurement framework for process evaluation in coding agents and instantiate step-level causal attribution with SCAE, a replay-based estimator derived from a structural causal model of agent execution. Our framework combines prefix-conditioned identification, replay/intervention-based estimation, and controlled judge-information manipulation to study process evaluation at the action, task, and step levels. Experiments on 499 file-localization episodes from 12 repositories show that next actions are driven primarily by execution provenance rather than code-graph transitions, execution uncertainty is structured at the task rather than step level, and full-trace judges exhibit systematic collider bias, suggesting that current process evaluation often measures semantic relevance rather than certified causal contribution.

While the top model on OmniDocBench now reaches 96.34% overall on printed-document parsing, the ability of current models to handle challenging handwritten documents remains largely uncharacterized. Existing benchmarks focus on isolated text or formulas, overlook handwritten tables and real-world degradation, and report aggregate accuracy without explaining why models fail. We present WildHandBench, a benchmark containing 500 handwritten documents across three structures (free text, tables, formulas), four languages, and nine real-world scenarios. We introduce a Prior-Driven Error (PDE) metric that quantifies whether errors originate from language priors rather than visual evidence. Evaluating 18 state-of-the-art models together with calibrated human baselines, we find: (1) the best model achieves only 71.85% overall; (2) humans outperform all models yet the gap is narrow (77.09% vs. 71.85%); and (3) model errors are qualitatively different from human errors – 63-91% of model errors are prior-driven versus only 49% for humans, exposing systematic reliance on language priors that conventional accuracy metrics cannot capture.

Concept-bottleneck controllable generation routes multi-attribute control through a low-dimensional concept code that, at deployment, must be synthesised from a target attribute configuration. We study this problem in concept-bottleneck text generation under multi-axis compositional generalisation, comparing three ways to obtain the inference-time code: classifier inversion against the encoder heads, reference-text encoding, and a post-hoc label-conditioned prior. Since a concept code admits no direct LM-fluency term, regularising inversion must instead constrain the code toward the encoder's training distribution. We therefore test bare inversion and three regularised variants: label-agnostic and label-conditioned Mahalanobis penalties, and a conditional normalising-flow density baseline. Every inversion variant we test underperforms a simple post-hoc prior fitted to per-combination encoder means on the same checkpoints, across three backbone families spanning $124$M to $8$B parameters. The bare form of classifier inversion also silently collapses to chance, traceable to a directly measured off-manifold code. We validate this diagnosis on real-world benchmarks and under external evaluators, enabling fair comparison with published baselines.

Symbolic protocol verification models the network attacker as a Dolev–Yao (DY) intruder, which does everything its knowledge permits, whether or not it serves any purpose; real adversaries instead maximise utility, attacking only when the payoff is positive. We introduce a rational Dolev–Yao attacker, a DY intruder whose actions carry costs and whose security-violating goals carry rewards, and call a protocol rationally secure when no intruder strategy achieves a violation with strictly positive utility, expressed in a weighted fragment of ATL (WATL). We prove this decidable for a bounded rational DY intruder over a finite cost-annotated concurrent game structure, characterise its complexity, and show it strictly refines DY security: some protocols are DY-insecure yet rationally secure, separated by a computable threshold. We illustrate the framework on two contrasting use-cases: an authenticated payment under session uncertainty, where a rational intruder must strategise across indistinguishable sessions and its imperfect information strictly raises the attack cost a designer must price against; and ThreeBallot, a cryptography-free scheme where we pinpoint the bribe-to-benefit ratio below which no rational coercer attacks.

We analyze a variant of stochastic gradient descent with initial regularization (SGDIR) and derive dimension-free upper bounds on its expected excess risk for the squared loss. In the noiseless case, we obtain new bounds for both averaged and non-averaged SGDIR under moment, source, and capacity assumptions. For a particular value of the source parameter, these bounds are of order $m^{-2}\log^{2}m$, where the number of training samples is of order $m$. For another value of the source parameter, we obtain, for any $\epsilon>0$, bounds of order $m^{-3+\epsilon}$, provided that the capacity parameter exceeds $\epsilon^{-1}$. We also establish a lower bound that matches our upper bounds in certain regimes up to a polylogarithmic factor. In the noisy case, we provide an instance-based comparison between SGDIR and ridge regression. Under general assumptions and a mild lower bound on the regularization parameter, we show that the expected excess risk of SGDIR is no larger than that of ridge regression, up to a polylogarithmic factor. Numerical experiments on synthetic and real data are consistent with our theoretical findings.

Floating waste in inland waterways threatens aquatic ecosystems and requires timely monitoring under cluttered, multi-object conditions. Existing aquatic-waste datasets provide limited geographic coverage, sparse multi-instance annotations, and little supervision beyond boxes and labels. Compact vision-language models (VLMs) therefore remain insufficiently evaluated for jointly localizing, classifying, counting, and explaining floating waste. We introduce WADE, a reasoning-annotated benchmark containing 2,167 images from rural Bangladesh, 13,608 bounding boxes, and ten waste categories. Each annotation is associated with class-level recognition rules covering visual cues, likely confusions, and discriminative features. We evaluate six VLMs under zero-shot, two-shot, reasoning-guided, and fine-tuned settings using detection, counting, and hallucination metrics. For resource-efficient adaptation, we jointly fine-tune Qwen3-VL-2B on boxes, labels, and reasoning chains using QLoRA. Fine-tuning increases recall from 0.0248 to 0.2339 and F1 from 0.0257 to 0.2163, while reducing image-level hallucination from 0.6836 to 0.0883. However, over three-quarters of instances remain undetected, establishing WADE as a challenging benchmark for dense floating-waste grounding with compact VLMs.

Large language models are increasingly used to propose research ideas, yet the prevailing ways of judging such ideas supply no shared decision rule: free-form judging sways with style and position, and scoring against a later paper rewards recovery of one realized trajectory. We introduce a benchmark that carries a proposal from Literature to Test: the Lit2Test benchmark centers on a six-field contract organized around a falsifying outcome, so that every proposal precommits the observation that would prove it wrong, making its quality decidable in the first place rather than merely arguable. Built prospectively from 200 real-paper neighborhoods, Lit2Test elicits proposals from four frontier models and compares them through 1,200 pairwise comparisons judged blind in both presentation orders. The protocol audits its own reliability through diagnostic controls and bounded human calibration, with three annotators corroborating the conclusions within explicitly stated reliability bounds. Lit2Test recovers a strict ranking of the four models in all 10,000 bootstrap replicates, and the separation comes from the quality of the proposed tests and metrics rather than from surface fluency. We release the benchmark, construction pipeline, and audit artifacts for public use.

Accurate simulation of three-dimensional time-harmonic wave propagation over many wavelengths requires control of phase error and efficient solution of large indefinite systems. We develop a three-grid solver based on the compact 27-point interpolated optimized finite-difference (IOFD) discretization. Its wavenumber-dependent stencil supports a fine-grid resolution of six points per shortest wavelength and an unshifted physical correction on the \(2h\) grid at only three points per shortest wavelength. The method retains unshifted IOFD operators on the \(h\) and \(2h\) grids, while a complex-shifted \(2h\)–\(4h\) auxiliary cycle preconditions a factorization-free iterative approximation of the coarse inverse. Restricting the shift to this auxiliary cycle preserves the propagative character of the coarse correction. Comparison with the outgoing Green function confirms phase and relative-amplitude accuracy on a sequence of meshes up to \revision{\(10240^3\)}, \revision{exceeding one trillion unknowns,} with the largest problem spanning approximately \revision{1704} wavelengths per coordinate. The same fixed solver configuration retains robust convergence across smooth, discontinuous, high-contrast, and geophysical velocity models and exhibits scalable parallel performance. In particular, a \revision{\(2560^3\)} problem spanning approximately \revision{425} wavelengths in each coordinate direction is solved in \revision{43.8 seconds} on just 64 NVIDIA A100 GPUs.

Federated learning has traditionally been formulated as a single-objective optimization problem, primarily focused on maximizing model utility. In real-world applications, however, machine learning models often need to optimize multiple and potentially conflicting objectives simultaneously. This motivates federated multiobjective optimization (FMOO), which provides a natural framework for jointly handling multiple task-specific objectives in federated learning. In this paper, we propose a momentum-based variance-reduced algorithm for federated multiobjective optimization. The method incorporates a momentum-driven gradient estimator into the local updates to reduce the variance of stochastic updates, leading to an improved convergence rate. We establish theoretical guarantees showing that the expected Pareto stationarity measure of a randomly selected output iterate decays at a rate of $\mathcal{O}(T^{-2/3})$, improving upon the $\mathcal{O}(T^{-1/2})$ rates established for existing methods such as FSMGDA and FedCMOO. Numerical experiments on federated multiobjective optimization benchmarks demonstrate the effectiveness and competitive performance of the proposed algorithm.

Conventional network threat detection primarily relies on packet-level, flow-level, or host-level telemetry. This paper investigates a different observation surface: unintended electromagnetic(EM) emissions generated by network interface card(NIC) activity, and asks whether such physical leakage contains sufficiently structured information for network threat-behavior recognition. We present NICWhisper, which externally captures NIC EM emissions, transforms raw measurements into time-frequency representations, and recognizes network behaviors without inspecting packet contents or host-side runtime states. Rather than competing with traffic-based detection, NICWhisper exploits the physical manifestation of traffic-driven NIC activity, whose timing, rate, concurrency, and burst organization naturally shape the measured EM leakage. We construct a NIC EM dataset covering active benign workloads and seven representative threat behaviors under diverse execution conditions, and systematically evaluate signal dependence, execution variation, measurement perturbation, and cross-device transfer. NICWhisper achieves 80.67% Macro-F1 across eight behavior classes, while further experiments show that the observed behavior-related information extends beyond simple signal magnitude and remains partially transferable across execution conditions and NIC hardware. These results establish NIC EM leakage as a complementary physical observation source for network security monitoring when direct access to conventional traffic or host telemetry is limited or undesirable.

Large Language Models (LLMs) have demonstrated remarkable capabilities in code generation, yet their performance degrades significantly on low-resource Hardware Description Languages such as Verilog. While multi-candidate sampling improves the likelihood of generating correct solutions, au-tomatically selecting the optimal candidate remains an open challenge. Through a systematic empirical study across nine models and two benchmarks, we identify two critical limitations:(1) existing execution-based reranking methods, which rely on testbench pass/fail outcomes, exhibit poor domain transferability due to low-quality generated testbenches; and (2) LLM-as-a-Judge suffers from reasoning hallucination, producing incon-sistent judgments for execution-equivalent code. These findings reveal two signal types with orthogonal errors: execution signals(deterministic but testbench coverage limited)and reasoning signals (semantically rich but hallucination-prone). Their orthog-onality suggests combining the two signals, yet in our experiments letting the reasoner directly observe execution results merely anchors its judgments on test outcomes; we therefore acquire the two signals independently and fuse them only at the decision stage. Based on these insights, we propose EAHC, an Execution-Anchored Hallucination Calibration reranking framework that anchors reasoning judgments to execution behavior so that execution-equivalent candidates receive consistent scores, which implements a dual-channel architecture: EAHC-R, a 4B reasoning discriminator; and EAHC-T, a testbench generator leveraging RAG for execution verification.

Quality control during printed circuit board (PCB) assembly is a critical step in ensuring reliable electronic products. Detecting misaligned pins during or after pin insertion remains a particularly challenging inspection task. This paper presents an automated defect detection method for identifying incorrectly inserted pins on PCBs. The proposed pipeline combines semantic segmentation using a U-Net architecture with contour-based feature extraction and logistic regression for board-level pass/fail classification. Segmentation masks are used to derive contour representations of individual pins, from which board-level features -such as average contour size- are extracted and used to train a logistic regression classifier. We evaluate the method on two datasets: an industrial collection of real-world PCB images, and a publicly available PCB pin-inspection dataset with substantially different visual characteristics. To assess the effectiveness of the proposed approach, a comparison against PatchCore, an anomaly detection technique new to be applied to pin inspection, as well as instance segmentation-based pin detection is made. The developed method achieved Area Under the Receiver Operating Characteristic Curve (ROC-AUC) values of 0.990 on a random test set split from the industrial data and 1.000 on the public dataset indicating strong separation between pass and fail boards. The results indicate that the proposed approach is a promising candidate for automated pin inspection in industrial environments and achieves strong performance on datasets with substantially different visual characteristics after dataset-specific training.

Agentic AI is a fascinating new tool for software development. It is a huge step forward compared to "conventional" AI assisted coding, which in turn was a considerable breakthrough earlier. AI support through LLMs is a young and very fast-moving field. The "conventional" (non-agentic) flavor became useful and productive in early 2025 (around 18 months ago) and the agentic flavor followed in fall 2025 (approximately 9 months ago). Besides all its benefits and potential, it also carries some fundamental risks for IT security. And the agentic approach added very severe risks while making others much more dangerous. With all the motivation to explore this fascinating new tool we should not ignore the risks but actively address them. We present (I) an assessment of the IT security risks, (II) a concept for mitigating them without breaking its benefits, and (III) an overview about an implementation of our concept. In this very dynamic field this is likely not the final and once-and-for-all answer to the identified issues but still a substantial step forward in responsible usage of Agentic AI for software development. It should also be a contribution to the community to allow early and eager evaluation of the potential of agentic AI for software development without actually suffering from its implied IT security risks.

Agent runtimes can Checkpoint an execution, Fork it, Restore a checkpoint, or Merge branches without restarting a task. We call these operations execution edits, with Checkpoint recording the current execution for later use and Fork, Restore, and Merge changing what the Agent will do next. An execution edit cannot undo an earlier authorization or a tool request already sent. An unsafe edit can therefore authorize the same tool action twice, discard a result the task still requires, or conflict with a call that began before the edit. The Agent is untrusted, so the runtime uses its execution record to determine which past actions an edit must account for and which required results it must preserve to keep the subsequent execution safe. Yet existing Agent systems support such operations without deriving what each edit must preserve from the running execution, whereas prior methods for computing safe behavior take that requirement as input. We give an algorithm that decides exactly whether an edit is safe. It returns all safe ways to continue, or proves that none exists. To make this decision, the algorithm lists every way the task can finish without violating policy. It removes any way that could make a still-required result impossible to finish later. If none remain, it returns a checkable proof that no safe implementation exists. Otherwise, the remaining ways describe exactly what the runtime may allow. Our formal results cover Checkpoint and the six forms of Fork, Restore, and Merge, together with extensions, atomic enforcement, and the information every exact checker needs. Lean mechanizes the finite checker and runtime invariant, and tests validate all six edit forms. The source code, Lean proofs, and executable tests are available in the public GitHub repository at https://github.com/eunomia-bpf/agent-check-restore-safety.

Gaze is increasingly used as an input signal for vision and multimodal models, yet no consensus exists on how to represent it across datasets. Raw traces preserve detail but are noisy and device-dependent, while coarse event labels are easy to model but can discard local motion structure. We formulate event-aligned, fixed-horizon angular displacement as an interpretable, event-conditioned motion vocabulary and compare it with event-only, spatial, absolute-angle, learned vector-quantized, and continuous representations. To assess transfer alongside target predictability and token collapse, our evaluation combines next-token prediction with target-domain regret, low-order target references, paired bootstrap, order sensitivity, motif overlap, and frozen structural probes. In an event-aligned headset benchmark, angular-motion tokens have lower target-domain regret than frozen-codebook VQ tokens in one transfer direction, while the reverse direction is inconclusive. The probes reveal complementary representation properties, and event-only tokens show that low perplexity can retain little motion information. On a third egocentric dataset, a matched comparison of I-VT, native, and frame-span interfaces shows that event construction materially changes transfer: native events have the lowest regret into EGTEA, while frame-span events have zero motif overlap and fail severely as a source. Motion-based tokenization therefore provides a compact representation for event-aligned egocentric gaze streams, while the evaluation identifies how target predictability and event construction shape cross-dataset conclusions.

Quantum computers pose a fundamental threat to blockchain systems that rely on elliptic-curve cryptography. This work reviews the quantum vulnerabilities and associated economic risks of major blockchain platforms, with a focus on Bitcoin, Ethereum, and Solana. We distinguish between at-rest, on-spend, and on-setup attacks and identify the blockchain components most exposed to quantum adversaries. We further review practical migration strategies toward post-quantum security, including NIST-standardized digital signatures and emerging solutions for Solana, Algorand, and Ethereum.

Dense IoT networks require reliable communication despite limited spectrum and substantial multi-user interference while maintaining manageable receiver complexity. This work introduces a deep-learning-based end-to-end multi-user communication design for interference-limited finite-blocklength IoT scenarios, focusing on short and medium blocklengths. We extend a prior 2-user SiameseNet transceiver framework to accommodate 2, 4, and 8 users, leveraging learned redundancy for interference suppression and noise robustness. Compared to conventional non-orthogonal access baselines, our method demonstrates strong Block Error Rate (BLER) performance across various scenarios without resorting to joint detection; the per-user decoder scales roughly linearly with the number of users. Further, we examine the robustness under interference mismatch and unequal interference strengths, critical for practical deployments with heterogeneous devices. The Latent-space analysis reveals that the learned codeword distance increases as the effective per-user rate decreases, corroborating with the observed BLER improvements. In addition, we also present preliminary results for a 2X2 MIMO setup under fixed-channel CSIT and CSIR, indicating potential for extending the framework to IoT gateways with multiple antennas.

We present HelaBERT, a family of two BERT-based masked language models pre-trained from scratch on approximately 1 billion tokens of Sinhala text sourced from MADLAD-400, CulturaX, and a custom corpus comprising news articles, Sinhala Wikipedia, and web crawl data. HelaBERT-Small (~23.3M parameters, 6 layers) and HelaBERT-Large (~110M parameters, 12 layers) both use a SentencePiece Unigram tokenizer (vocabulary size 32,000) tailored to Sinhala's agglutinative morphology and complex script. We evaluate both models on four downstream Sinhala text classification tasks: news category classification, news source classification, sentiment analysis, and writing style classification, using 5 independent seed runs with stratified 80/20 train/test splits. We additionally propose a dual pooling classification head and evaluate it systematically across all four tasks, finding consistent improvements on sentiment analysis and a moderate gain on news category classification for HelaBERT-Small, while the standard [CLS]-linear head remains competitive on news source classification, a headline-level task with short average input length. We release both models to support further research in Sinhala NLP.

Multi-behavior recommendation (MBR) leverages auxiliary behavioral signals, such as clicks and add-to-cart, to enhance target behavior prediction like purchases. While recent graph neural network-based approaches have achieved strong performance by systematically propagating auxiliary behavior signals, they still suffer from two fundamental challenges inherent to auxiliary behaviors: (1) missing auxiliary signals, which hinder generalization to items without auxiliary observations, and (2) unreliable auxiliary signals, which amplify noise misaligned with the target behavior. To address these challenges in a unified manner, we propose BOAR, an environment-conditioned MBR framework that addresses missing and unreliable auxiliary signals through two complementary modules conditioned on auxiliary observability. Extensive experiments demonstrate that BOAR consistently outperforms state-of-the-art baselines, achieving up to 7.82% gains in HR@10 overall and up to 44.2% gains for target items without auxiliary observations, highlighting its ability to capture hidden preferences beyond observed auxiliary relations. Our code is available at: https://github.com/LSH0411/BOAR.

For next-generation air interfaces, two-sided artificial intelligence (AI) models have received growing attention, with AI models deployed at both the transmitter and receiver for efficient channel feedback and data communication. However, their practical deployment is complicated by assumptions commonly made in existing studies, including isolation from legacy users, training under predefined channel conditions, and gradient-based fine-tuning requiring substantial cross-vendor communication. This article revisits these assumptions and presents practical alternatives. First, for legacy coexistence, we integrate two-sided model processing into the 5G New Radio (NR) protocol stack and validate its operation alongside conventional NR on a real-world testbed. Second, instead of training under a massive number of predefined channel conditions, we construct a compact model table by jointly optimizing two-sided models with trainable surrogate channels, and select the best model according to the current channel condition to enable channel adaptation with high task performance and low training/storage overhead. Finally, unlike existing fine-tuning that exchanges large gradient vectors containing potentially private model information, we present gradient-free zeroth-order fine-tuning that requires only scalar feedback, facilitating multi-vendor interoperability. Together, these approaches advance the practical deployment of two-sided AI models while highlighting key open challenges.

We present TSWAP, a deployed eight-language conversational wellness advisor grounded, via retrieval-augmented generation, in a verified knowledge base of Thai traditional medicine and certified wellness providers. An unmodified open-weight LLM (Qwen3.6-35B-A3B on vLLM) is grounded on a ~30.6K-chunk Thai index by a hybrid dense-sparse retriever with cross-encoder reranking; a first-turn query classifier forces tool-based retrieval for entity lookups; a rule-based safety layer enforces medical scope and Thai emergency routing; and all eight languages are served zero-shot with translate-then-retrieve. We release the first Thai traditional-medicine/wellness retrieval benchmark (50 questions with gold document IDs; Recall@5 = 0.88), production QA logs (91.1% test-retest pass over 259 cases), and a 71-question frontier no-retrieval probe showing what each grounding pillar contributes: without the safety prompt the backend model family produced a full drug-dosing schedule and complied with out-of-scope requests, and without the knowledge base it produced zero verifiable provider recommendations. We further report two transferable deployment findings: English-calibrated 4-bit AWQ quantization corrupts Thai tone marks, and forced-retrieval routing is necessary for reliable grounding.

Vision-language models are known to encode spatial information in their hidden states, yet often fail to use it when answering. However, it remains unclear when and where this encoded information reaches the answer. We address this with direction patching, a class-conditioned causal intervention applied across layers, token positions, and prompt formats. Using spatial-ID directions constructed following prior encoding evidence, we find that causal influence on answer logits emerges only at mid-to-deep depths. Text chain-of-thought suppresses immediate object-word argmax-level transport in most models, while visually grounded prompts keep it open. Positive target-logit gain can remain below the argmax threshold, and transport can re-emerge at the final prefix token or at the answer step in deeper layers. Across the ten VLMs we study, these local effects form descriptive transport patterns. Complementary experiments characterize how these patterns shift across datasets, attributes, and encoding amplitudes. Together, these results reframe the encoding-grounding gap as a problem of conditional transport in VLMs.

Inference-time pipelines often sample multiple outputs, filter them with a learned safety model, and return the proxy-feasible output with the highest learned reward. We show that this composition creates a two-stage failure: an imperfect safety proxy first contaminates the feasible set with unsafe outputs, and reward maximization can then amplify this residual contamination. We define safety hacking as selecting an output that passes the learned constraint but violates the true safety criterion. For constrained Best-of-$N$ sampling, we derive finite-$N$ bounds governed by the joint upper reward tails of safe and unsafe outputs within the proxy-feasible set. If unsafe-but-feasible outputs have the heavier tail, safety hacking becomes asymptotically certain as $N$ grows, even when false-positive mass and average safety- and reward-proxy errors are arbitrarily small. We also show that policies within a bounded $\chi^2$ divergence from the proxy-feasible reference distribution admit an $N$-independent safety-hacking bound, and instantiate this general coverage-control principle with constrained pessimistic sampling. Coverage control limits amplification but cannot repair a contaminated feasible set: admitted unsafe outputs may still be favored, and regularized selection is not necessarily safer than constrained Best-of-$N$ for every reward proxy. Toy and language-model experiments characterize both contamination and its reward-tail amplification, which exposes an inherent difficulty in inference-time scaling with learned safety models.

This report summarizes the contributions and results of the 1st Asynchronous CASTLE Challenge at the Joint Egocentric Vision Workshop in conjunction with CVPR 2026.

We develop a finite-width geometric framework describing how learned feature geometries are organized, transported, and selectively aligned in deep neural networks. Incompatibility among weight-generated covariance, gates, and backward sensitivities is quantified through three families of commutators: between gates and covariance, between sensitivities and covariance, and between average gradient outer products (AGOPs) and neural feature matrices (NFMs). An exact layerwise identity decomposes the sensitivity-covariance commutator into four sources: downstream transport, adjacent-layer imbalance, pointwise sensitivity fluctuations, and nonlinear gate-covariance interactions. The AGOP-NFM commutator is a singular-value-weighted transport of the internal commutator, explaining why observed feature-side alignment alone does not determine the internal geometry from which it emerges. Buffered localized energies resolve mixing between separated covariance subspaces. We establish spectral-gap, projector-evolution, and stabilization estimates, and formulate conditional Lyapunov principles that yield decay under explicit geometric error-bound or intrinsic-damping assumptions. These criteria do not follow from gradient flow alone and clarify why risk reduction need not imply commutator collapse. Analytic examples and numerical experiments exhibit factorization of spectral and activation geometry, transient growth, and cancellation among nonzero sources. In tested finite-time regimes, cancellation dominated by a negative transport-imbalance interaction persists across depths, widths, and two regression benchmarks. Spectral alignment therefore appears as a layer- and scale-dependent compatibility phenomenon governed by transport, interaction, cancellation, and possible damping, rather than a universal consequence of training.