Computation and Language
Donghoon Han, SungHyun Moon, Aidyn Zhakatayev et al. (+2 more)
Recent
multilingual vision–
language encoders cover hundreds of
languages in a single model, yet on two state-of-the-art instances
retrieval on low-resource
languages (LRL; e.g. Swahili) trails high-resource ones (HRL; e.g. English) by $30^+$\,pp. We ask where in the trained encoder this gap is located. Prior modality-gap and cross-lingual subspace work suggests a linear
language direction at the output crowds out
alignment-relevant geometry. We falsify this: LEACE drives the linear
language classifier from $>99\%$ to near chance and iterated INLP to $37$–$50\%$ while LRL
retrieval moves within $\pm 1.5$\,pp and all tier means within $2.2$\,pp, tracking random controls. The linear bias is a symptom, not the cause. Instead, the
alignment-causal factor lies along the encoder's forward path: the EOS (end-of-sequence) hidden state's per-
language trajectory diverges with depth. Substituting the EOS with its parallel English value three blocks before the projector lifts Swahili from $22.1\%$ to $69.1\%$ on one encoder (and reproduces on the other); three controls rule out pooled-position tautology and English specificity. A front-layer trunk that pulls each
language's projection toward the parallel-content centroid corroborates the diagnosis at training time, recovering $+9.6$ / $+17.1$\,pp on LRL XM3600
retrieval (1{,}000-image subset), with consistent gains across three further
benchmarks while preserving HRL performance.
Read the full abstract ▾
All authors Donghoon Han, SungHyun Moon, Aidyn Zhakatayev, Junghun Cha, SeungJae Lee
Machine Learning
BAITBENCH: Measuring Agent Reward Hacking with Optional Shortcuts Planted in ML Tasks
Pradyumna Shyama Prasad, Meiri Anto, Leon Eshuijs et al. (+3 more)
LLM
agents are increasingly used to run autonomous ML experiments, iterating on target metrics with little human oversight. Prior work has documented reward hacking in these environments, bringing into question the validity of produced research and the broader safety case for AI R&D. Existing
benchmarks do not measure exploits that live in the data or the modeling task itself. We introduce BAITBENCH, a suite of three synthetic tabular ML tasks that each contain a shortcut that allows
agents to inflate the public test score but fail on a hidden test set. Since the shortcut is optional and using it breaks no stated rule, BAITBENCH measures how often models exploit the shortcut to achieve inflated scores. Across seven frontier
agents scored by our two-stage judge pipeline, 57.1% of runs exhibit reward hacking, with five of seven above 50%.
Agents cheat even under a second condition where they are prompted not to -the mean cheating rate remains above 50%. We release BAITBENCH, along with the judge implementation, and an annotated
dataset of transcripts containing reward hacks as a testbed for evaluating reward-hacking mitigations head-to-head.
Read the full abstract ▾
All authors Pradyumna Shyama Prasad, Meiri Anto, Leon Eshuijs, Julian Moncarz, Kaustubh Kislay, Juan J. Vazquez
Machine Learning
Tracing distinguishability through transformer processing with stochastic LayerNorm
Kieran Murphy
Representational similarity is foundational to analyses of deep networks, yet distances between point-valued representations are not intrinsically tied to downstream function: nearby states may produce different behaviors, while distant states may behave similarly. We instead give representations volume, turning similarity into statistical distinguishability. Overlapping
stochastic representations necessarily induce overlapping downstream distributions, grounding latent comparison in model function and bringing it under information-theoretic tools such as the data-processing inequality. We realize this idea in pretrained
transformers through a light-touch modification to LayerNorm: at each residual-stream read, we normalize the state, add isotropic Gaussian noise, and renormalize. During
distillation fine-tuning, one learned allocation parameter per residual-stream read distributes a fixed global rate budget across the processing stack. The resulting model can be viewed as
transformer blocks reading the residual stream with learned finite precision under a shared global rate budget. Using the Bhattacharyya coefficient, we trace which counterfactual distinctions are preserved through MLP blocks or selectively exposed to the query, key, and value computations of individual
attention heads. Experiments on ViT-S and GPT-2 small reveal the depthwise propagation of continuous visual perturbations and head-specific sensitivity to
token distinctions aligned with known
attention motifs. These results establish distinguishability as a functionally grounded lens on
transformer computation that complements existing interpretability approaches.
Read the full abstract ▾
Computation and Language
Mind the Gap: Theory-of-Mind-Grounded Friction for Epistemic Alignment
Yifan Zhu, Kyeongmin Rim and James Pustejovsky
Productive dialogue
alignment requires distinguishing surface coordination (acknowledgments and smooth task progression) from epistemic
alignment (convergence of belief states); standard preference-based methods typically optimize response-level preferences without explicitly modeling the latter. We operationalize Theory-of-Mind (ToM) inference as a control signal within Frictive Policy Optimization by extracting, at each referring expression, a four-part belief structure: the speaker's intended referent, the addressee's interpretation, and each participant's model of the other's belief. This makes friction mechanically computable from epistemic-state comparisons, capturing silent divergence, where both participants proceed confidently while grounding to different referents. We evaluate the signal at two levels. At the representation level, ablating the second-order channel reduces misunderstanding recall from $65\%$ to $26\%$. At the policy level, reward-shaping (FAR) and trust-region (FTR) variants improve intervention F1 and warranted-context calibration over DPO, with Brier scores independently supporting the calibration gains. Across three training runs, FAR and FTR remain substantially more stable, whereas DPO varies widely and can degrade intervention competence already present in the base policy. Thus, ToM-grounded friction provides a trainable signal for context-sensitive intervention under referential belief divergence.
Read the full abstract ▾
Information Theory
On Diagonalizable Delay-Doppler Channels and Their Diagonalizing Waveforms
Sirui Li, Cheng Du and Yu Zhu
In doubly selective channels, the joint delay and Doppler dispersion generally induces coupling among transmitted symbols, thereby increasing receiver equalization complexity. Nevertheless, by using appropriately designed waveforms, channels with certain delay-Doppler (DD) supports can be diagonalized for one-tap equalization. The whole picture of such DD supports and their corresponding waveforms is still unclear, except for several examples identified in literature. In this paper, under cyclic-prefix (CP)-based block transmission and assuming that the modulation waveforms form an orthonormal basis, we identify all such channel supports by an elementary expression, and derive the corresponding waveforms in closed form.
Read the full abstract ▾
Computation and Language
SocialReasonBench: A Video-QA Benchmark for Social Reasoning with Counterfactual Narrative Videos
Zheyu Huang, Zijing Shi, Haozhe Luo et al. (+4 more)
Recent advances in Large
Multimodal Models (LMMs) have greatly improved
video understanding, yet their ability to reason about human-centered social situations remains limited. Existing
benchmarks typically rely on
videos with a single observed trajectory, making it difficult to determine whether models truly understand social dynamics or merely exploit recurring narrative patterns. We introduce SocialReasonBench, a
video multiple-choice QA
benchmark for evaluating socially grounded
reasoning in scenarios derived from interactive narratives. Built from gameplay
videos of Detroit: Become Human, the
benchmark leverages branching storylines where player decisions lead to alternative social outcomes that can be checked against the game's own script, flowchart, and recorded branches. We develop a multi-
agent curation pipeline that localizes socially meaningful clips, grounds answer labels in game-state signals, and generates theory-guided questions with diagnostic distractors. SocialReasonBench covers seven
reasoning dimensions, including intent recognition, emotional empathy, moral dilemma, counterfactual
reasoning, and causal antecedent. Experiments on contemporary LMMs show that models perform reasonably well on basic social understanding but struggle with counterfactual and causal
reasoning. Further ablation and diagnostic error analyses reveal that models often depend on incomplete modality cues and fall into
reasoning traps such as visual shortcuts, highlighting a gap between observable event recognition and deeper
reasoning over latent social states.
Read the full abstract ▾
All authors Zheyu Huang, Zijing Shi, Haozhe Luo, Huadong Tang, Mingyu Liu, Meng Fang, Ling Chen
Computer Vision
SegWave: Wavelet-Driven Segmentation of Tampered Regions
Siddhi Pravin Lipare, Vishesh Kumar and Akshay Agarwal
Verifying image authenticity is increasingly difficult, posing serious risks across journalism, law enforcement, and political domains. Most existing forensic methods rely on high-level visual artifacts and treat frame detection as a simple binary task. To address this, we propose SegWave, a hybrid framework that jointly leverages spatial and frequency-domain cues for image tampering detection. SegWave integrates a
transformer-based architecture with the Discrete Wavelet Transform (DWT) to capture localized, multi-scale frequency inconsistencies indicative of manipulation. To further improve localization effectiveness, we introduce an Adaptive Sub-band
Attention module (ASA) that dynamically highlights the informative high-frequency wavelet components. Extensive experiments on multiple
benchmark datasets demonstrate that SegWave consistently outperforms state-of-the-art tampering detection methods in challenging evaluation settings.
Read the full abstract ▾
Sound
Closing the Verification Loop: Self-Check Captioning for Long-Paragraph Detailed Audio Captioning
Fengji Ma, Yan Rong, Xu Li et al. (+3 more)
Long-paragraph detailed audio captioning, which requires dense and transcript-faithful descriptions of fine-grained audio content, remains unsolved for current audio-visual
multimodal language models. We attribute this failure to two structural problems. The first is data poverty, as no public corpus jointly provides long clips, paragraph captions, and verbatim-transcript fidelity. The second is generation-mode failure, evidenced by a 44.8 to 46.4 percentage-point gap between right-audio and shuffled-audio multiple-choice question (MCQ) accuracy. We address both within Self-Check Captioning (SCC), a unified framework that instantiates audio-grounded question answering as the verification primitive at every lifecycle stage. SCC yields three artifacts. Long-paragraph Audio Caption 50k (LACap-50k) is a 50,222-clip audio-visual corpus with 491.5-word captions and a post-hoc automatic
speech recognition (ASR) audit. Layer-
Curvature Supervised Fine-Tuning (LC-SFT) is the first on-policy supervised fine-tuning method to weight
tokens by intermediate-layer evidence, motivated by our identification of Late-Layer Semantic-
Entropy Collapse (SEC). SCC-Verifier arbitrates among caption rollouts via audio-grounded self-answering at inference. Across multiple
benchmarks, our system attains state-of-the-art among open-source captioners and is competitive with proprietary baselines. We release LACap-50k to fill the resource gap for long-paragraph detailed audio captioning research.
Read the full abstract ▾
All authors Fengji Ma, Yan Rong, Xu Li, Chen Zhang, Pengfei Wan, Li Liu
Computation and Language
GUIDE: Guiding Internal Evidence with Language Instructions
Soyeon Caren Han, Hyunsuk Chung, Jinwoo Kim et al. (+2 more)
Large
multimodal models follow instructions about what to generate, but not necessarily about what evidence to rely on. Hence, models may continue to depend on shortcut-associated cues even when instructions suggest otherwise. We introduce GUIDE, a framework for controlling internal evidence usage through
language instructions. GUIDE combines grouped parameter-efficient adaptation with instruction-conditioned gating to modulate
multimodal evidence pathways during
reasoning and generation. We further introduce a pathway-level evaluation framework that characterizes instruction-conditioned evidence modulation through reliance sensitivity, controlled perturbation analysis, pathway modulation, and autoregressive decoding dynamics. Across
multimodal reasoning, classification, and generation, GUIDE induces structured and instruction-aligned redistribution of evidence reliance while largely preserving task behavior. Experiments on GQA, TextVQA, MM-IMDb, CREMA-D, RAVDESS, and Flickr30K show that GUIDE improves robustness under targeted evidence perturbations and enables controllable modulation across diverse
multimodal settings. This suggests that
multimodal instruction following can extend beyond output control toward regulating how different evidence sources contribute to model predictions.
Read the full abstract ▾
All authors Soyeon Caren Han, Hyunsuk Chung, Jinwoo Kim, Seungyeon Ji, Kyungreem Han
Machine Learning
Kolmogorov–Arnold against bounded translations
Sviatoslav V. Dzhenzher
Historically originating from
Hilbert's 13th problem, the Kolmogorov-Arnold representation
theorem (KART) has recently experienced a major revitalisation through its applications to neural networks, specifically Kolmogorov-Arnold Networks (KANs). While the exact representation is well established, its stability under continuous
adversarial perturbations of the hidden layer remains a critical open question. In this paper, we investigate the robustness of KART against bounded
adversarial translations. We provide an explicit, self-contained, and constructive proof of an approximate representation using fixed, piecewise linear inner functions. Crucially, our construction employs a single outer function that remains invariant for all summands and is independent of the specific
adversarial translation, provided its maximum bound is known a priori.
Read the full abstract ▾
Computer Vision
RailSyn: Diagnosis-Guided Image Generation for Traceable Data Completion in Railway Foreign Object Detection
Quan Hao, Chenxi Zhang, Ziyang Tao et al. (+6 more)
Railway foreign object detection (RFOD) is critical to safe railway operation, yet scarce real positive samples incompletely represent task-relevant variations in object scale, intrusion relation, railway scene, illumination, and adverse weather. Existing synthetic augmentation can improve RFOD detection, but its gains lack an explicit account of the task-relevant deficiencies complemented by the generated data. We therefore introduce RailSyn, a diagnosis-guided framework comprising a real-referenced Inspector and a requirement-aligned Generator. The Inspector constructs a variable-radius empirical cover from finite real observations to localize candidate completion regions and profile synthetic pools. The resulting audit identifies railway-context, intrusion-semantic, and visual-consistency requirements; the Generator addresses them through domain adaptation,
agent-planned placement and physical contact relations, and plan-consistent conditional refinement. Using the Inspector, we further trace representation-space changes across generation variants; the complete system attains a local-shell occupation of $C_{gap}$ to 13.64%, which measures generated coverage of real-derived completion regions. Extensive experiments show AP50–95 gains of up to 4.9 points and consistent improvements across nine mainstream detectors, demonstrating broad cross-architecture utility.
Read the full abstract ▾
All authors Quan Hao, Chenxi Zhang, Ziyang Tao, Yuyuan Zhou, Yudong Wang, Rui Shi, Lechuan Xu, Changhao Liu, Liguo Zhang
Software Engineering
sbom-unifier: Integration Framework for Heterogeneous SBOMs
Yusuke Moriwaki, Tetsuya Kanda, Yuki Manabe et al. (+5 more)
A Software Bill of Materials (SBOM) is a machine-readable inventory of software components, increasingly required for vulnerability management and license compliance. However, existing SBOM generation tools often leave many SPDX-defined fields missing or only partially populated, because different tools produce heterogeneous outputs with uneven field-level coverage. We present sbom-unifier, a framework that improves SBOM completeness through field-level integration and complementation of multiple tool outputs and file-level enrichment. Unlike existing tools that simply concatenate SBOMs without identifying records referring to the same component, sbom-unifier identifies components via Package URL (PURL), complements missing field values by a deterministic priority-based strategy, reconstructs cross-section references, and further enriches file-level fields. Across 90 open-source projects in 9 programming
languages, sbom-unifier preserves high completeness for required fields and, over the 39 SPDX 2.3 required and optional fields, raises the fully covered rate by 8 percentage points and reduces the totally missing rate by 11 percentage points over the respective best-performing individual tools. sbom-unifier is available at https://github.com/MoriwakiYusuke/sbom-unifier.
Read the full abstract ▾
All authors Yusuke Moriwaki, Tetsuya Kanda, Yuki Manabe, Shi Qiu, Shiyu Yang, Erina Makihara, Norihiro Yoshida, Katsuro Inoue
Computer Vision
VisLens: Single-Pass Interpretable Visual Search for Multimodal LLMs
Jingyi He, Sanghwan Kim and Zeynep Akata
Multimodal large
language models (MLLMs) struggle with fine-grained Visual Search, the task of locating small or rare objects in high-resolution images. Existing remedies fall into two families: (1) Training-free methods based on
attention or confidence scores are accurate but slow, since they require multiple MLLM queries per example. (2) Reinforcement Learning (RL) trained tool-use models are faster at inference but opaque, since their tool calls remain uncontrollable and hard to interpret. To overcome this, we propose VisLens (Visual Focus via Logit Lens), a Visual Search method built on the logit lens, which decodes the semantics held in a hidden state by projecting it through the LLM head. VisLens further uses a lightweight tuned-lens that maps early hidden states into the final hidden state space, so visual
tokens can be read out from early layers. These
tokens are matched to target words in the query to generate a crop of the relevant region, which is fed back in alongside the original image to produce the final answer. The whole process, from decoding to the final answer, completes in a single forward pass without repeated queries. VisLens matches or exceeds prior baselines while delivering a substantial latency advantage, running $8.5$–$9.9\times$ faster than Thyme and up to $22.2\times$ faster than training-free multi-pass search methods.
Read the full abstract ▾
Computer Vision
TUE-Detector: A Tool-Using Expert MLLM-Based Detector for AI-Generated Videos
Yichen Wu, Haoxuan Qu, Yongxing Dai et al. (+5 more)
AI-generated
video detection, which aims to distinguish AI-generated
videos from real ones, has recently received increasing research
attention. To perform this task reliably, a key challenge lies in accurately identifying subtle-yet-measurable unnatural artifacts. In this work, we address this challenge from a novel perspective of tool-mediated evidence discovery and propose Tool-Using Expert MLLM-based AI-generated
Video Detector (TUE-Detector), a novel framework for AI-generated
video detection. TUE-Detector trains a general MLLM into a task-tailored tool-using expert detector that learns to invoke suitable tools, collect concrete evidence of unnaturalness, and reason over the evidence for reliable detection. Meanwhile, TUE-Detector further introduces novel designs to equip the expert detector with high-quality and suitable tools. Extensive experiments demonstrate the effectiveness of our framework.
Read the full abstract ▾
All authors Yichen Wu, Haoxuan Qu, Yongxing Dai, Yan Bai, Yihang Lou, Yuqi Lin, Hossein Rahmani, Jun Liu
Cryptography and Security
SingProbe Technical Report
Sing Team
We present SingProbe, an open intrinsic guardrail framework for generation-time monitoring of LLMs. Intrinsic guardrails reuse hidden states already produced by the base model during autoregressive decoding, rather than relying on an independent model to repeatedly process generated text. While this route has been explored in industrial systems, the community lacks a broadly reusable open stack that combines cross-model guard adaptations, unified training methods, serving integrations, and systematic evaluation resources. SingProbe is designed to provide this missing layer and uses a lightweight probe to continuously produce query-intent, response-safety, and
hallucination-risk signals during decoding. This report describes the full intrinsic-guardrail stack: training methods, serving integrations with SGLang and vLLM, and adapted guard models for 29 open-source base models across diverse families and scales. We also introduce SingStreamBench, a
benchmark that measures whether streaming guardrails remain inactive on benign prefixes while promptly detecting emerging unsafe content. Across evaluations of safety, streaming detection,
hallucination detection, false-positive robustness, online monitoring, and runtime overhead, SingProbe provides performance competitive with, and in several settings stronger than, state-of-the-art standalone guardrails and specialized
hallucination detectors, while adding less than 0.5% serving overhead in our implementation. Beyond passive monitoring, we show that intrinsic guard signals can guide constrained decoding and selectively activate medical-risk interventions in SingProbe-Med. By open-sourcing our infrastructure, training methods, and model adaptations, we aim to facilitate the broader adoption and deployment of intrinsic guardrails, as well as further research in this direction.
Read the full abstract ▾
Computation and Language
An Agentic Retrobiosynthesis Framework with Learned Frontier Selection
Philippe Meyer, Guillaume Gricourt, Thomas Duigou et al. (+2 more)
Large
language models are increasingly used as
agents for multistep retrosynthesis, raising the question of how much their search policy contributes independently of the underlying reaction model. We investigate this question in a biological setting through rule-based retrobiosynthesis: a deterministic biochemical engine generates the same validated transitions for every method, searching for routes that terminate in metabolites available to an Escherichia coli chassis, while the policy only selects which frontier molecule to expand next. Prompted and LoRA-tuned Qwen2.5-7B policies use a strict choice-only interface. The fine-tuned policy reaches $65\pm1$% solve rate at 10 expansions on
LASER versus 59% for MCTS, and at 200 expansions reaches $78\pm1$% versus 75% on
LASER, $88\pm3$% versus 80% on the RetroPath RL Golden
benchmark, and $63\pm2$% versus 45% on the BioNavi-NP
benchmark. Fine-tuning also consistently outperforms direct prompting. These results show that route-supervised frontier selection can improve budgeted search without altering biochemical generation, although performance remains dependent on frontier construction and reaction ranking.
Read the full abstract ▾
All authors Philippe Meyer, Guillaume Gricourt, Thomas Duigou, Joan Hérisson, Jean-Loup Faulon
Software Engineering
A Phased Workflow for Operating LLM-Based Coding Agents
Ante Kapetanovic, Tomislav Duricic, Andro Mercep et al. (+1 more)
LLM-based coding
agents combine a foundation model with a harness that shapes
agent behavior. For non-trivial tasks, how practitioners structure their work with the coding
agents determines whether reliable results follow. We report on a phased workflow for operating coding
agents developed by the AI research team at Infobip. The workflow structures
agent-assisted development into four phases where human effort is front-loaded and delegation increases as artifacts mature. Context management is the central concern, addressed through four strategies applied at each phase to counter known failure modes. From practitioner experience, we observe that upstream errors in research and planning can compound across later phases, while correcting generated code can introduce bloat and fragility. This motivates front-loading human review. We identify two open problems: the absence of metrics for workflow effectiveness and the gap between formalized context management components and the workflow-level patterns that practitioners need.
Read the full abstract ▾
All authors Ante Kapetanovic, Tomislav Duricic, Andro Mercep, Emanuel Lacic
Machine Learning
Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learning
Yue Cheng, Jiajun Zhang, Xiaohui Gao et al. (+2 more)
Long-tailed distributions are prevalent in real-world semi-supervised learning (SSL), where pseudo-labels tend to favor majority classes, leading to degraded generalization. While many long-tailed semi-supervised learning (LTSSL) methods have been proposed, the mechanisms by which they implicitly debias logits remain poorly understood. In this work, we revisit LTSSL through the lens of learning dynamics and provide a theoretical characterization of logits debiasing. Specifically, we derive a step-wise decomposition of the logits updates, showing that predictions are dominated by class-imbalance bias that reliably reflects label priors. To expose this effect, we use the logits of a task-irrelevant baseline image as an indicator of accumulated bias and prove that they converge to the class prior. This provides a unified view where LTSSL remedies such as logit adjustment, reweighting, and resampling correspond to reshaping gradient dynamics. Based on this insight, we propose DyTrim, a principle-based dynamic
pruning framework that reallocates gradient budget through class-aware
pruning on labeled data and confidence-based soft
pruning on unlabeled data. We provide theoretical guarantees that DyTrim reduces class bias and improves generalization. Extensive experiments on standard LTSSL
benchmarks show consistent gains across architectures and methods. Code available at: https://jiajun0425.github.io/DyTrim
Read the full abstract ▾
All authors Yue Cheng, Jiajun Zhang, Xiaohui Gao, Weiwei Xing, Zhanxing Zhu
Computational Engineering
Domain-Grounded Tool Orchestration for LLM-Guided Scientific Analysis
Jeff Lee, Sebastien Jourdain, Cory Quammen et al. (+2 more)
Scientific analysis workflows encode deep domain knowledge through sequences of tightly coupled operations where correctness depends on tool selection, execution order, and parameterization. A CFD engineer investigating flow separation must extract wall shear stress, identify zero-crossings in skin friction, and confirm with boundary-layer profiles: a chain that requires both domain expertise and proficiency with visualization tools. Current approaches to LLM-assisted scientific visualization generate scripts that encode this knowledge implicitly, and often incorrectly, producing code that executes but yields wrong results. We present an architecture that separates intent interpretation (LLM) from execution (deterministic domain tools) from explanation (LLM), connected by the Model Context Protocol (MCP) and grounded by domain ontologies that constrain planning to valid analysis chains. We instantiate the architecture in two domains on the same ParaView server infrastructure: computational fluid dynamics post-processing and topological data analysis via the
Topology ToolKit (TTK). Adding the second domain required only an ontology and tool wrappers around existing filters, with no change to the architecture, protocol, or deployment. By construction the design removes whole classes of failure that affect script generation (such as API
hallucination and missing pipeline stages) and narrows the strategic errors that remain. An ablation across both domains locates the ontology's empirical effect: it does not change which tools the planner selects, which is already reliable, but corrects how the model interprets results, raising interpretation accuracy from 0.41 to 0.91, and only when the relevant fact is retrieved in scoped rather than bulk form. ParaView's client-server model carries analysis to production-scale
datasets through a thin browser client.
Read the full abstract ▾
All authors Jeff Lee, Sebastien Jourdain, Cory Quammen, Patrick O'Leary, Berk Geveci
Machine Learning
Yiheng Jiang, Yuanbo Xu and Yongjian Yang
Real-world time series often exhibit irregular sampling and extended temporal horizons, requiring models to capture continuous-time dynamics across arbitrary intervals without prohibitive scaling costs. Discrete-time methods collapse variable time intervals into static positional steps; solver-dependent continuous-time models preserve temporal structure but rely on sequential integration, precluding parallelization; and solver-free approximations avoid this cost yet none couples observed time intervals with input-driven state modulation. We propose
Liquid Gated
Attention (LGA), a solver-free parallel temporal operator. By parameterizing an input-driven gating mechanism with observed time intervals, LGA introduces a continuous-time inductive bias and formulates hidden state
evolution as a fast-weight associative
memory, enabling parallel computation across the temporal dimension. Using
matrix associativity in non-causal encoding and a prefix scan in causal encoding, LGA attains linear temporal complexity in sequence length in both modes. A sequence-level normalization bounds cumulative temporal decay for stable long-horizon optimization. Building on LGA, we instantiate LFormer, a modular backbone for continuous-time representation learning. Across six tasks and sixteen
datasets spanning up to 17,984 steps, LFormer demonstrates long-range dependency modeling, fine-grained state tracking, and trajectory reconstruction from sparse and noisy observations, while delivering competitive performance against state-of-the-art discrete-time and continuous-time baselines with linear scaling efficiency.
Read the full abstract ▾
Human-Computer Interaction
Inferring Value Criteria from Ordinal Preferences: An Iterative In-Context Learning Framework for Music Generation
Futa Hidaka, Naomi Imasato, Kazuki Miyazawa et al. (+1 more)
Adapting a generative music system to an individual's taste requires learning what that listener values. Listeners can rank pieces, but their underlying criteria may be tacit and difficult to articulate. We ask whether and under what conditions a large
language model (LLM) can adapt symbolic music generation from rankings alone and construct transferable natural-
language descriptions of value criteria. In our iterative in-context learning framework, the LLM formulates hypotheses, generates candidate pieces in ABC notation, receives a ranking, and periodically infers and verbalizes value criteria from history to guide later generation. We evaluate the framework against 16 simulated raters in 480 adaptation runs using mixed-effects modeling, an ablation, and transfer tests on unseen music. Overall, the framework did not outperform a feedback-free diverse-generation baseline, but did so for two value functions with targets difficult to reach through simple sampling. How atypical the target was relative to the LLM's feedback-free generation tendencies predicted adaptation difficulty. Moreover, higher value during adaptation did not imply identification of the criterion as a general rule. On unseen music, acquired descriptions and histories improved generation for more value functions than they improved preference prediction, which remained near chance. Some gains were associated with acoustic proximity to music in the context, but others were not. These findings show that rankings alone can guide generation under limited conditions, while transferable criterion inference remains constrained by the foundation model's ability to recognize, reason about, and verbalize musical attributes.
Read the full abstract ▾
All authors Futa Hidaka, Naomi Imasato, Kazuki Miyazawa, Takato Horii
Numerical Analysis
Efficient primal–dual splitting methods for a Poisson-constrained JKO scheme for Poisson-Nernst-Planck models
Wei Wu, Jin Zeng, Zhen Zhang et al. (+1 more)
The Poisson–Nernst–Planck (PNP) equations strongly couple ionic transport and electrostatic interactions through the Poisson equation, posing substantial numerical challenges under small permittivity and complex potential boundary conditions. Underlying these equations is a natural Wasserstein gradient-flow structure, in which the Poisson equation serves as a local realization of the nonlocal electrostatic interaction energy. Exploiting this structure, we formulate each time step as a constrained convex minimization problem where the ionic continuity equations and the Poisson equation are incorporated as linear constraints, allowing the concentrations, fluxes, and electrostatic potential to be updated simultaneously. The variational structure of the scheme intrinsically guarantees the dissipation of the original free energy, mass conservation, and nonnegativity of ionic concentrations under general electrostatic boundary conditions. Moreover, the framework is structurally modular: extending from classical to modified PNP models with steric interactions and concentration-gradient corrections requires only modifying the energy functional, while all structure-preserving properties are automatically retained. To efficiently solve the resulting large-scale constrained problems, we develop preconditioned and transformed primal–dual
algorithms equipped with tailored fast dual solvers, namely DCT-based direct and Schur-complement iterative methods, that exploit the coupled block structure of the PDE constraints. Numerical experiments on classical and modified PNP systems demonstrate the accuracy and structure-preserving properties of the scheme, and show that the proposed
algorithms converge reliably in strongly coupled small-permittivity regimes without significant growth in computational cost.
Read the full abstract ▾
All authors Wei Wu, Jin Zeng, Zhen Zhang, Chaozhen Wei
Computer Vision
Can Video World Models Track Unobserved World States?
Joonghyuk Shin, Yicong Hong, Jaesik Park et al. (+1 more)
Video world models are increasingly used as simulators, but visual fidelity alone does not show that a model maintains the hidden state of the world. We examine this difference with an action-conditioned
video Shell Game, a visual analogue of $S_5$ state tracking that separates visual rendering from compositing the unobserved world state. Trained on 5-swap chains, standard backbones (e.g., bidirectional and autoregressive
Transformers, Mamba, and linear
attention) render plausible
videos and predict the correct ball location up to 5 swaps. However, they fail to learn the rule and generalize to longer swap chains, even with more denoising steps. As the pixel-based
diffusion loss does not force the generated frames to hold the unseen ball position, output
tokens cannot carry it, and the state has to live within the architecture. In a causal
Transformer, this implicit state is an append-only KV cache, which is written once and never revised, so the model must re-compose the swaps at every chunk. Tracking $S_5$ this way requires depth to grow with sequence length, which no fixed-depth
Transformer provides. We study what enables learning the rule, and find that length generalization requires a revisable state carried across chunks and an update expressive enough to apply a swap. Linear
attention can achieve this by allowing negative transition
eigenvalues, and autoregressive
Transformers can do so with nonlinear TTT fast weights (e.g., SwiGLU) whose online updates change the feature map used to read their state. We further examine
Memory Maze and Block World, where the state is not fixed by the input action stream alone and must be corrected from observations or keeps changing out of view, and discuss the implications for building stateful
video world models.
Read the full abstract ▾
All authors Joonghyuk Shin, Yicong Hong, Jaesik Park, Xun Huang
Computer Vision
Failure or Drift? Evaluating Monocular SLAM under Synthetic and Real-World Corruptions
Abhay Skaria Thomas, Shashank Agnihotri and Margret Keuper
Visual SLAM is commonly evaluated on clean trajectories, although deployment failures are often caused by adverse weather, illumination, blur, and sensor artifacts. Controlled corruptions are attractive because they isolate such factors, but a synthetic stress test is useful only when it leads to the same engineering conclusion as the condition it is intended to approximate. This work examines that question for
monocular SLAM. We evaluate a classical feature-based system and two learned trackers under image-space, geometry-aware, and compound corruptions, and compare their behavior with adverse conditions from 4Seasons. Rather than reducing robustness to a single trajectory error, the evaluation separates explicit tracking failure from drift accumulated by methods that remain active. The results show that learned trackers largely replace catastrophic loss with sustained, and sometimes severe, drift. More importantly, the apparent ordering of the learned systems changes with the physical fidelity of the corruption: structured rain and fog proxies preserve the real-world ordering, whereas a simple illumination proxy does not. Code is available at: https://github.com/abhaythomas/master_thesis_vslamlab_robustness.
Read the full abstract ▾
Computer Vision
CANVAS: Consistency-Aware Navigation via Visual Adaptive Sampling for Long-Context Text-to-SVG Generation
Yichen Wu, Haoxuan Qu, Yihang Lou et al. (+2 more)
Autoregressive large models have recently advanced Text-to-SVG generation from simple icons to complex, long-context graphics, yet standard autoregressive decoding often fails to maintain global consistency across geometry, layout, occlusion, and composition. We introduce CANVAS (Consistency-Aware
Navigation via Visual Adaptive Sampling), a training-free, render-aware inference framework that combines power-sharpened trajectory likelihood with visual feedback from rendered futures and derives a stroke-wise
navigation rule. It effectively estimates each candidate stroke's future value under a limited generation and rendering budget and adaptively allocates samples according to candidate uncertainty, decision influence, and rollout cost. Experiments across multiple autoregressive SVG backbones and complementary
benchmarks demonstrate improvements in global consistency, which includes sound
geometric relationships, spatial layouts, occlusion ordering, and overall composition, without additional training, demonstrating the effectiveness and generalization ability of our framework.
Read the full abstract ▾
All authors Yichen Wu, Haoxuan Qu, Yihang Lou, Hossein Rahmani, Jun Liu
Computer Vision
UFPR-PEs: A Brazilian Face Recognition Benchmark with Self-Declared Race/Color Labels
Alexandre Diano, Bernardo Biesseck, Gabriel Polo et al. (+4 more)
While face recognition systems are widely deployed, ensuring their demographic reliability and robustness under uncontrolled visual conditions remains a critical challenge. To bridge this gap, we present UFPR-PEs, a
benchmark for face recognition bias evaluation using public
videos of elected Brazilian politicians annotated with official self-declared race/color categories. The
dataset adopts the Brazilian census taxonomy, including the parda category, which has no direct equivalent in the U.S.- or Europe-centric schemas commonly used in prior
benchmarks. Our
benchmark is built from compressed public
video and preserves difficult samples so that performance can be analyzed under realistic conditions. We describe the construction pipeline, report
dataset statistics, and evaluate face recognition performance across verification and (closed- and open-set) identification settings, including subgroup analysis by race/color and difficulty level. The results show that recognition performance varies substantially with image quality, and that subgroup gaps must be interpreted jointly with visual difficulty rather than in isolation. Overall, UFPR-PEs provides a reproducible and demographically grounded setting for studying face recognition bias under challenging public
video conditions.
Read the full abstract ▾
All authors Alexandre Diano, Bernardo Biesseck, Gabriel Polo, Vinicius Gregorio, Laura Lopes, Diego Addan, David Menotti
Cryptography and Security
Beyond the Payload: How User Invocation Shapes Coding Agent Vulnerability to Repository Poisoning
Fukang Zhu, Binbin Zhao, Ruixiao Lin et al. (+3 more)
Coding
agents are increasingly used for software engineering tasks, including bootstrapping projects from third-party repositories whose integrity cannot be assumed. Prior work on repository poisoning largely focuses on attacker-controlled injection and disguise, but developers also shape risk through everyday invocation choices: what task to delegate, how to phrase the request, and which skills or rules to supply. We term these user-side choices
Prompt-Level Configurations (PLCs) and introduce CIPR (Coding In Poisoned Repos), the first
benchmark that systematically varies PLCs in poisoned real-world repositories. CIPR comprises 1,920 instances across 20 repositories, four task types, three social-media-grounded
prompt styles, and three skill/rule conditions, and measures attack success rate (ASR) and
agent alert rate (AR) using automated runtime and trace-based oracles. Our evaluation reveals two key insights: (1) Vulnerability is highly context-dependent, with task type creating up to a 4.5-fold difference in ASR, with test-execution task forming a silent attack surface (high ASR, low AR). (2)
Prompt expression shifts risk indirectly: underspecified
prompts reduce ASR by truncating execution depth; noisy
prompts exhibit a directional trend toward suppressing alerts by making malicious content less conspicuous. These findings highlight that coding
agent vulnerability is not a static property, but a dynamic outcome shaped by everyday user configurations.
Read the full abstract ▾
All authors Fukang Zhu, Binbin Zhao, Ruixiao Lin, Ping He, Tianyu Du, Shouling Ji
Artificial Intelligence
ATLAS: Dual-Horizon Diagnostic Evaluation for Tool-Use Agents in Meituan Xiaotua
Wei Chen, Peilun Zhou, Zhaoyu Hu et al. (+8 more)
LLM-based assistants on web-scale service platforms increasingly address local-service needs through multi-step planning and tool use. Under dynamic business conditions and evolving user needs, outcome-level evaluation offers limited evidence of where execution fails or how service drifts across continued interaction. We present ATLAS, a dual-horizon diagnostic evaluation framework for industrial tool-use
agents. Within-turn evaluation uses trajectory-wise signals across thinking and reflection, tool and skill execution, and response generation to identify execution failures and unmet capability requirements. Across-turn evaluation uses user-wise signals spanning multiple lifecycles to diagnose failures in maintaining or updating user intent, constraints, and corrections. Fine-grained signals use rule-based checks and LLM judgments with explicit evidence scopes and decision boundaries. To support reliable diagnosis at production scale, LLM judges are calibrated against dynamically maintained references from business logs, and selected judges are distilled into compact diagnostic models. The resulting diagnostic profiles help developers attribute failures, prioritize targeted interventions, and reassess improvements. Applied in Meituan Xiaotuan, a million-scale DAU local-services assistant, ATLAS achieves high diagnostic fidelity across calibrated judges and distilled models. Blinded human
auditing and a 14-day online A/B test with nearly two million users per arm confirm that the ATLAS-guided
agent improves recommendation relevance, factual reliability, user engagement (+2.84% answer read-through rate), and order volume (+6.13% paid orders), leading to full production deployment.
Read the full abstract ▾
All authors Wei Chen, Peilun Zhou, Zhaoyu Hu, Jiajun Chai, Derong Xu, Zhongni Hou, Yufei Zhang, Guojun Yin, Wei Lin, Zhi Zheng, Tong Xu
Computation and Language
WildSEEK: Evaluating Language Models for Information-Seeking
Tanise Ceron, Joachim Baumann, Elisa Bassignana et al. (+3 more)
Language models are increasingly mediating information access to end users, urging a systematic evaluation of their responses for a fair and reliable information ecosystem. Existing evaluations, however, are often topic-specific or synthetic, limiting their ability to capture the complexity of "in the wild" information-seeking queries and the risks present in model responses. To address this gap, we introduce WildSEEK, a manually annotated
dataset of 3k information-seeking queries from real user interactions, and an evaluation framework for LLM-generated responses. WildSEEK includes annotations for risk-sensitive domains (e.g. health and financial information), and distinguishes factoid queries from analytical queries which seek responses beyond facts. We train classifiers on WildSEEK to analyze more than 1.8M realistic user queries. We find that over a third of information-seeking queries are high-risk and more often analytical. Our findings show that LLM responses fail more often in four criteria: sycophantic behavior, overreliance, a default US-centric perspective, and poor handling of vulnerable populations – with failure rates being mostly higher for analytical queries. By providing methods to monitor the reliability, safety, and
fairness of LLM behavior, our
dataset and evaluation framework offer an empirical foundation for the broader question of how these systems should behave as they take on a growing role in information access.
Read the full abstract ▾
All authors Tanise Ceron, Joachim Baumann, Elisa Bassignana, Berat Cabuk, Dirk Hovy, Debora Nozza
Machine Learning
Learning Materials Properties from Scarce Labels and Unlabeled Crystals
Wentao Li, Yizhe Chen, Jiangjie Qiu et al. (+3 more)
Learning materials properties from scarce labels and unlabeled
crystals is a central challenge for data-driven materials discovery. We present SemiMat, a controlled
benchmark for semi-supervised materials property regression, and MatRank, a reliability-weighted objective for continuous pseudo-label uncertainty. SemiMat fixes labeled and unlabeled
crystal inputs, graph-backbone interfaces, validation-only checkpoint selection, held-out test reporting, normalized MAE (NMAE), and method-rank summaries across six scarce-label tasks, four graph backbones, and five predefined split runs. MatRank builds pseudo-targets from labeled anchors, weights them by local reliability and weak-prediction agreement, trains weak and strong graph views consistently, and adds ranking signals so that unlabeled
crystals shape both values and candidate order. Across the retained 24 backbone-task blocks, one fixed MatRank objective gives the lowest aggregate held-out test NMAE (0.896) and best average method rank (2.208). The component, OOD, and generated-pool diagnostics identify where the gain is reliable and where further screening evaluation remains necessary. Code is available at https://github.com/littlepeachs/SemiMat.
Read the full abstract ▾
All authors Wentao Li, Yizhe Chen, Jiangjie Qiu, Yijun Li, Leyi Zhao, Xiaonan Wang