Named after the hundred-eyed watchman of Greek myth, Argus watches the education landscape: spotting new opportunities, pressure-testing the ventures we're building, and tracing every read back to the real-world signals behind it.
The evidence library: the raw signals the pipeline is watching across the education ecosystem. Every idea is built from these.
arXiv:2607.00664v1 Announce Type: new Abstract: We propose YOMI-Bench, a benchmark for evaluating kanji reading and phonological understanding of large language models (LLMs) for Japanese. In Japanese, a single kanji character often has multiple possible readings, making it difficult to infer the correct reading from surface-level text alone. Due to these linguistic characteristics, it is empirically known that LLMs exhibit low performance in kanji reading for Japanese. The proposed YOMI-Bench consists of four tasks specifically designed to evaluate kanji reading performance in Japanese. In our evaluation using YOMI-Bench, we assessed one multilingual open LLM, four Japanese-specific open LLMs, and five commercial LLMs. As a result, we found that even Japanese-specific models show low performance, and that commercial models also perform poorly on generation tasks that require consideration of kanji readings.
arXiv:2607.00661v1 Announce Type: new Abstract: Explanations for emotion classifiers are usually produced post hoc, with no guarantee that they reflect the computation behind the label. We present an explication interface for event-based emotion analysis. A parser maps the input text to an explication, a short script in the closed vocabulary of Natural Semantic Metalanguage organized into twelve typed slots, and a fixed decision list of rules transcribed from published semantic definitions computes the label from the explication alone. The faithfulness guarantee is therefore causal and definitional, while all empirical risk lives in the learned parser, which the per-line entailment interface makes auditable against the input. On crowd-sourced event descriptions, our fine-tuned parser reaches 0.33 accuracy and 0.48 selective accuracy on a small held-out set, suggesting that the interface trades insignificant accuracy difference to a black-box model for a verifiable, inspectable decision
arXiv:2607.00605v1 Announce Type: new Abstract: Limited Memory Language Models (LMLMs) externalize factual knowledge to a database to enable deletion-based unlearning without retraining. Existing evaluations measure post-deletion correctness in aggregate and cannot tell whether a deleted fact persists through residual parametric memory, alternative retrieval paths, or near-neighbor retrieval artifacts. We propose a causal auditing framework that holds the model fixed and varies the database state at inference time across three interventions: FULL, DEL-ON, and DEL-OFF. The framework decomposes post-deletion behavior into parametric leakage L(f), retrieval-mediated correctness R(f), and a retrieval artifact rate grounded in the inference-time retrieval trace. We apply it to 12,228 alias-closure deletions across thirteen databases, including four adversarial topologies (Base, Alias, Noise, Collision) we construct in three domains, and six prompt formulations. Parametric leakage is near ze
arXiv:2607.00601v1 Announce Type: new Abstract: The game of Taboo requires describing a target word without using a set of forbidden words, so that other players can guess it. This deceptively simple task combines strict lexical constraints with the need for communicatively effective descriptions, making it a compelling playground for examining how LLMs navigate competing demands at inference time. We evaluate two open-weight models under conditions that intervene at progressively deeper levels of the generative process, from prompting to generation-time constraints to internal representations manipulations. We assess their outputs through forbidden word violation detection, LLM-as-a-judge measuring the degree to which generated descriptions successfully evoke the target concept for both human and machine guessers, and examining whether the strategies models adopt under constraint align with those of human players. Our results show that compliance with the rules of the game and communi
arXiv:2607.00597v1 Announce Type: new Abstract: Scientific literature search often requires more than retrieving papers from a single query: users' intents are underspecified, preference-dependent, and evolve through interaction. Existing search agents typically rely on fixed pipelines or implicit language-only reasoning, making their search strategies difficult to control, inspect, and refine. We introduce PaperPilot, a multi-turn literature search agent that frames scientific search as workflow induction. Given an anchor paper and a user query, PaperPilot constructs an executable DAG of paper-search operators, including keyword search, citation expansion, filtering, scoring, reranking, and evidence extraction. User feedback is then used to refine both the query and the workflow itself. We train PaperPilot with supervised workflow imitation and preference optimization over controlled workflow corruptions. Experiments show that PaperPilot-9B improves over the base Qwen3.5-9B toolset ag
arXiv:2607.00588v1 Announce Type: new Abstract: Continuous diffusion language models such as ELF report record-low generative perplexity (Gen-PPL). We find a catch: these models repeat far more than human text, and Gen-PPL rewards rather than penalizes that repetition, so its low scores overstate quality. Strip the repetition and ELF-B's Gen-PPL rises from $19.5$ to $27.7$; the smallest model even posts the best Gen-PPL because it repeats most. We trace the repetition to its source: a contractive attractor along a \emph{single direction} in the self-conditioning feedback loop, the loop that feeds each step's clean estimate into the next. Because the failure is one-dimensional, a one-dimensional fix suffices, and we propose one. \textbf{ACE} (Attractor-Contrast-Escape) subtracts that single, label-free direction from the feedback at each step. Estimated once on the $105$M model, the direction cuts repetition to near the human level while keeping quality competitive, and transfers near-u
arXiv:2607.00576v1 Announce Type: new Abstract: Multi-image content has become an increasingly prevalent form of visual communication in social media, giving rise to a new safety issue, multi-image implicit toxicity (MIIT), where each image appears benign in isolation, but harmful semantics emerge when the images are interpreted jointly. MIIT is particularly challenging for existing commercial moderation APIs and models due to the lack of explicit risky cues in each image. This paper aims to study how to identify MIIT. We first provide a formal definition of MIIT and analyze three key challenges for its detection. To alleviate the scarcity of data in this area, we construct MIIT-dataset, an image-only multi-image safety dataset covering seven representative risk categories through an automatic generation pipeline. Finally, we train MiShield with progressively distilled reasoning supervision, enabling it to produce safety judgments accompanied by explicit analyses of the correlated enti
arXiv:2607.00570v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) increasingly requires models to answer questions from multiple retrieved documents, where only some sources are relevant and the retrieved bundle may contain stale, noisy, or conflicting evidence. Existing contrastive decoding methods primarily focus on resolving conflicts between the model's internal memory and the retrieved context. In contrast, we study the complementary problem of intra-context conflict in multi-document RAG. To evaluate this setting, we introduce DRQA, a factual-conflict question answering benchmark derived from enterprise deep-research scenarios, where answers are grounded in synthetic enterprise-specific facts that are designed not to be recoverable from the model's internal memory. We further propose Dual-Confidence Contrastive Decoding (DCCD), a training-free decoding method that combines document-level confidence, which estimates whether a document appears sufficient for answ
arXiv:2607.00502v1 Announce Type: new Abstract: While long-horizon mobile GUI agents typically rely on thought-action-observation loops, they struggle to separate persistent task states from transient screen observations. As execution histories grow, this entanglement imposes a severe context burden, causing agents to forget initial requirements, hallucinate progress, or repeatedly interact with stale interfaces. To address this, we introduce Task-State Representation (TSR), a training-free framework that explicitly decouples task state from sensory input. Acting as a lightweight external wrapper, TSR maintains three structured components: a global instruction summary, a dynamic progress tracker for subgoals, and a transition-aware action verifier. By continuously updating through pre- and post-action visual comparisons, TSR effectively guides the agent's reasoning without requiring architectural modifications. Experiments across four mobile GUI benchmarks validate TSR's effectiveness,
arXiv:2607.00501v1 Announce Type: new Abstract: We present BaseRT, a native Metal inference runtime for large language models (LLMs) on Apple Silicon, and report the highest inference throughput on this hardware to date. Existing runtimes, including llama.cpp and MLX-based frameworks, incur overhead from abstractions not designed for Metal's execution model or Apple Silicon's unified memory topology. By building natively on Metal with chip-specific kernel fusion, unified memory-aware optimisation, and custom dispatch logic, BaseRT recovers performance that framework-based approaches leave on the table. BaseRT supports a wide range of model families across eight quantisation formats (Q2 to FP16) on all Apple M-series devices. In this paper, we evaluate the Qwen3, Llama 3.2, and Gemma 4 families at Q4 and Q8 quantisation on M3 and M4 Pro devices. BaseRT achieves up to 1.56x higher decode throughput than llama.cpp and up to 1.35x higher than MLX, with substantially larger margins on prefi
arXiv:2607.00485v1 Announce Type: new Abstract: Large reasoning models (LRMs) have achieved strong reasoning capabilities in English, yet their performance degrades significantly when required to reason in other languages. A natural solution is to transfer the model's English reasoning ability to target languages. However, existing transfer approaches typically rely on distilled target-language reasoning traces from stronger LRMs or online supervision from external judge models, which are costly and difficult to scale. In this paper, we propose PCS (Progressive Code-Switching), a more efficient transfer framework that requires only lightweight translation without any stronger model for distillation or judging. PCS first constructs code-switched reasoning traces by translating a subset of English reasoning steps into the target language, and uses them to initialize the model's code-switching ability via supervised fine-tuning. It then applies reinforcement learning with a step-level lan
arXiv:2607.00482v1 Announce Type: new Abstract: Reasoning language models frequently overthink: generating extended chains of behaviors such as hedging, approach abandonment, and self contradiction that consume tokens without improving answers. We show that these behaviors are not merely a consequence of length; even when controlling for response length, incorrect traces exhibit higher rates of unproductive self-reflection than correct ones. Addressing this requires identifying where self-reflection helps vs hurts, but obtaining these step-level annotations is costly. We observe that intermediate answer commitments within reasoning traces can provide a cheap proxy: by comparing each final answer candidate in the trace to the ground truth, we can determine whether subsequent reflection is productive without any additional supervision. Building on this insight, we propose DASH (Drift Aware advantage SHaping), which assigns segment-level credit based on whether each reasoning segment lead
arXiv:2607.00447v1 Announce Type: new Abstract: Large language models often produce hallucinated answers that violate prompt-level constraints. A key diagnostic question is whether these failures reflect missing knowledge, or whether the model has the relevant information but follows the wrong inference path. We study this phenomenon as inference misalignment: a mismatch between the answer supported by the prompt and the answer favored by statistically salient latent associations. We formalize this view with a latent key-task model, in which pretraining-frequency imbalance can cause a shortcut path to dominate the constraint-sensitive path and induce positive inference loss. The framework predicts two failure modes: task-retrieval bias in entity disambiguation and key-selection bias in action choice. We introduce TrapQA, a controlled diagnostic testbed with two components. ScientistQA tests disambiguation among similar scientists with supplementary factual probes, while Real-Life Const
arXiv:2607.00423v1 Announce Type: new Abstract: Vision language models (VLMs) demonstrate strong zero-shot performance, but often perpetuate social stereotypes in person-centric queries, yielding skewed demographic distributions. Current debiasing methods apply uniform bias corrections across all input queries regardless of their bias sensitivity, creating a fundamental fairness--utility trade-off. Strong debiasing distorts semantically meaningful information in bias-insensitive queries, while weak debiasing fails to mitigate stereotypes in bias-sensitive ones. This one-size-fits-all approach hampers simultaneously achieving high utility on bias-insensitive queries and fairness on bias-sensitive queries. We introduce Reward-Gated Test-Time Adaptation (RG-TTA), a reinforcement learning-based test-time adaptation framework that selectively applies debiasing based on input sensitivity. RG-TTA adaptively triggers fairness regularization based on the bias sensitivity of each input during te
arXiv:2607.00418v1 Announce Type: new Abstract: This paper presents Speech Playground, an interactive speech visualization and comparison tool. While existing tools such as Praat are excellent, it can be cumbersome to integrate them with modern deep learning representations and use them for comparison. Speech Playground addresses this by combining a Python backend with a web-based frontend for interactive exploration of multiple feature types, including continuous, discrete, and variable-length representations. It includes TextGrid and forced alignment support together with configurable distance and alignment settings for visual and auditory comparison. Speech Playground is intended for use in speech research, representation validation, and computer-aided pronunciation training (CAPT)-oriented experimentation.
arXiv:2607.00415v1 Announce Type: new Abstract: Authority bias poses a critical safety concern in language models: models systematically prioritize social cues from authority figures over factual consistency, swaying their answers based on source credibility rather than evidence. We mechanistically investigate this phenomenon using a controlled medical QA setting, where hints suggesting incorrect answers are attributed to personas of varying expertise. Across Llama-3.1-8B, Qwen3-8B, and Gemma-2-9B, we find that models respond in a graded manner proportional to perceived authority, a hierarchy that is never explicitly prompted but emerges from training. Logit lens analysis and linear/non-linear probing localize this effect to a critical late layer where correct answer representations are actively erased, an erasure that scales with authority level, resists mean vector intervention, and is only partially reversible through chain-of-thought reasoning. Our findings suggest that authority-i
arXiv:2607.00368v1 Announce Type: new Abstract: Large language model test-time training (TTT) is often evaluated through local proxy metrics: models are updated on recent tokens, retrieved context, target-domain data, or verifiable task attempts, and then judged by perplexity, future-token loss, long-context performance, or reward. These metrics are well matched to claims about stream adaptation, domain adaptation, context compression, and reward-backed test-time improvement. They are weaker evidence, however, for a capability that TTT results are increasingly used to motivate: deployed assistant memory, personalization, or sparse post-deployment learning, which instead requires behavioral evidence such as later recall, paraphrase robustness, retention, locality, conflict handling, and use in downstream actions after the original support context is removed. We introduce a behavioral evaluation framework that calibrates TTT memory claims to the evidence that supports them. It has two co
arXiv:2607.00341v1 Announce Type: new Abstract: Large language models achieve strong performance on many reasoning tasks when allowed to externalize intermediate steps as Chain-of-Thought (CoT). However, many questions require the model to internalize the multi-step reasoning within a single forward pass before generating the answer. We study this challenge through two-hop reasoning, a representative task where the model must compose multiple pieces of parametric knowledge within a single forward pass. Standard non-recurrent Transformers suffer from a depth-local storage problem: facts learned in earlier layers are unavailable where second-hop retrieval happens. We found that Looped Transformers mitigate this issue by reusing the same memory, but still generalize imperfectly. We show that the remaining bottleneck is representational. In the two-hop reasoning task, the first loop often makes the correct bridge entity nearly perfectly decodable, yet the corresponding hidden state remains
arXiv:2607.00339v1 Announce Type: new Abstract: Conversational data is increasingly used as a persistent source of user state for long-running assistants and AI agents. However, querying this data remains challenging because conversations naturally evolve: plans are revised, preferences change, and later messages frequently supersede or contradict earlier information. Existing long-memory pipelines largely treat memories as independent text or vector objects. This approach often retrieves semantically similar but stale evidence, offering limited support for state-aware reasoning. To address this problem, we present TRACE, a query processing framework over temporal evidence graphs for evolving conversational data. TRACE models conversations as a hierarchical graph spanning events, sessions, and topics, enriched with typed temporal, causal, update, and contradiction relations. Crucially, the framework maintains validity annotations so obsolete facts remain accessible for historical queri
arXiv:2607.00274v1 Announce Type: new Abstract: Effective writing feedback is among the strongest drivers of student learning, yet producing it at scale is labor-intensive. LLMs offer a natural path to scaling writing support, but two gaps stand in the way: few public corpora capture how instructors actually deliver feedback in real classrooms, and no reliable method measures whether generated feedback aligns with what an instructor would write. We address both. SEFORA is a public corpus pairing instructor inline feedback with assignment prompts, rubrics, scores, and multi-draft revisions across various college writing genres, comprising 564 drafts and 8,240 instructor annotations. UniMatch is a reference-based evaluation framework for open-ended generation: it segments feedback into feedback units, scores their semantic correspondence under instructor-derived criteria, and aligns them via optimal matching to yield interpretable precision, recall, and F1. Across 74 experimental configu
arXiv:2607.00250v1 Announce Type: new Abstract: Maltese has decent text corpora and pretrained language models, but, like many languages outside the handful with large OCR benchmarks, only a single known real labelled PDF corpus for OCR training, 57 page, far below what paragraph-level training needs: low-resource for OCR specifically. With no real corpus to train on at scale, we built a synthetic training pipeline and a 5-stream Tesseract LV-ROVER ensemble, and report results on a 422-paragraph benchmark against a fine-tuned-Tesseract baseline of character error rate (CER) 0.0234. Ensemble recognition alone improves CER by 44 percent, to 0.01317; a five-stage post-processing chain brings the full pipeline to CER 0.00700, a 70 percent reduction. Most of that chain is typographic normalisation, but one stage recovers misread diacritics rather than aligning punctuation, so we report it as a recognition gain rather than folding the whole chain under one label. We treat the 44 percent figu
arXiv:2607.00208v1 Announce Type: new Abstract: Reinforcement learning for diffusion large language models (dLLMs) has largely moved to trajectory-aware methods. The current state of the art, TraceRL, holds that random masking is mismatched with the model's inference trajectory, and it reconstructs that trajectory during training by slicing each rollout into up to K/s trajectory-aligned training samples, a cost that grows with the block size K. We show that this mismatch can be mitigated without reconstructing the trajectory. Our method, SLIM-RL, bounds the commit risk of each rollout step with a tau-budget decoder, reducing aggregate commit risk in the training data. During optimization, SLIM-RL trains on these risk-controlled rollouts with a trace-free random-masking objective that adapts variance-reduction tools, combining sequence-level importance sampling, deterministic quadrature over masking levels under a mean-preserving, monotonically decreasing per-block mask schedule that we
arXiv:2607.00185v1 Announce Type: new Abstract: Khipus--knotted cord devices--were the primary recording medium of the Inka Empire (c. 1400-1532 CE), yet their system remains undeciphered. We present a reproducible machine-learning pipeline applied to the Open Khipu Repository (OKR), a public database of 619 khipus comprising 54,403 cords and 110,677 knots. We engineer 27 structural features per khipu and apply (i) unsupervised clustering via UMAP and HDBSCAN, recovering three structurally distinct groups (silhouette = 0.769); (ii) supervised provenance classification via gradient boosting, reaching F1 = 0.86 for the Inka Late Horizon imperial style; and (iii) SHAP-based interpretability, which identifies cord twist direction as the dominant structural discriminator of imperial khipus. We further report two findings of methodological interest. First, one cluster is dominated not by a geographic region but by nineteenth-century European museum collections, indicating that colonial acqui
arXiv:2607.00171v1 Announce Type: new Abstract: Text embeddings are standard for semantic similarity tasks, yet their evaluation remains an open challenge. Current benchmarks are static, cover only a limited set of languages, are often domain-specific, susceptible to overfitting, and poorly representative of low-resource languages. To address these limitations, we introduce ALEE, a framework that extends Sentence Smith (Li et al., 2025) to the cross-lingual and paragraph level. ALEE uses Abstract Meaning Representations (AMR) to generate English minimal pairs with controlled, fine-grained semantic shifts, which are paired with translations in target languages. This approach enables targeted diagnostics for models in any language with English parallel data. We conduct a large-scale empirical study across a diverse set of embedding models and 275+ languages spanning three parallel datasets. On ALEE, performance varies substantially across languages, text lengths, and linguistic phenomena
arXiv:2607.00159v1 Announce Type: new Abstract: Knowledge-Based Visual Question Answering (KB-VQA) aims to evaluate whether Visual Language Models (VLMs) can retrieve, ground, and reason over external structured knowledge beyond visual evidence. In practice, answer accuracy is widely adopted as the primary evaluation metric, implicitly treating correctness as a proxy for knowledge-grounded reasoning. However, for existing KB-VQA benchmarks, this proxy relies on critical assumptions that are often overlooked and rendered unreliable by benchmark issues: annotated answer must be derivable from the associated knowledge base, question must be well-posed with sufficient constraints, and visual setting must meaningfully require grounded disambiguation. In this work, we show that these assumptions are systematically violated in existing KB-VQA benchmarks. Our audit reveals substantial instances with missing or contradicted answers and underspecified questions that render accuracy a misleading
arXiv:2607.00158v1 Announce Type: new Abstract: Hallucination remains one of the central obstacles to deploying medical LLMs. Yet, even when hallucination can be detected, it is still unclear whether the internal representations associated with it can be used for control rather than detection alone. Using four open-source models across a suite of medical question-answering datasets, we show that a simple, carefully conditioned probe can reliably detect hallucination, with AUROC scores between 0.77 and 0.86 in our case. We further show that this signal is distributed and redundant rather than narrowly localized. Systematically selected neurons outperform random neurons only at very small subset sizes, whereas random subsets of a few hundred neurons recover nearly the full signal, and low-dimensional random projections preserve most of the detection performance. Beyond detection, we test whether this representation is causally actionable. Across 16 model--dataset combinations, our result
arXiv:2607.00143v1 Announce Type: new Abstract: Online hate speech has been linked to a global rise in violence against minorities, including incidents such as mass shootings, lynchings, and ethnic cleansing. Societies grappling with this issue, particularly when hate speech targets specific groups based on religion, race, ethnicity, culture, nationality, or migration status, face the challenge of balancing freedom of expression with the need for effective content moderation on widely used online platforms. In response to this challenge, we introduce a comprehensive hate speech dataset covering five distinct topics in Turkish: refugees, the Israel-Palestine conflict, anti-Greek sentiment in Turkey, ethnic or religious communities (Alevis, Armenians, Arabs, Jews, and Kurds), and LGBTI+, alongside one topic in Arabic (refugees). In addition, we develop state-of-the-art BERT-based models to address multiple dimensions of hate speech analysis, including hate category classification, hate i
arXiv:2607.00139v1 Announce Type: new Abstract: The cost of human expert evaluation is a principal bottleneck to deploying language models in specialized, high-stakes domains. This is particularly acute for Arabic sociolinguistic knowledge: credible grading requires not only linguistic fluency but deep cultural familiarity that cannot be approximated by surface-level metrics. We address this with a cross-evaluation framework instantiated on two underrepresented Arabic dialect communities: Egyptian and Iraqi Arabic. We contribute 103 validated prompt-rubric pairs (70 Egyptian, 33 Iraqi; 53 Cultural, 50 Linguistic), authored and graded by native-speaker SMEs using penalty-weighted rubrics distinguishing positive content requirements from answer-specific negative error criteria. Three frontier LLMs serve as target models (graded by human SMEs across 302 unique prompt-response pairs), while five frontier LLMs serve as automated judges enforcing a provider-level self-evaluation guard. A dua
arXiv:2607.00083v1 Announce Type: new Abstract: Language models have changed from unreliable text generators to highly-capable large models with trillions of parameters. Capability increases come hand-in-hand with increases in scale, making understanding the internal representations of models more challenging. Since millions of users increasing rely on language models to interact with external tools or make decisions in medium or high-stakes scenarios, we need to establish control over model behavior and know when to trust model outputs. In this paper, we discuss our contributions on harnessing the latent spaces by proposing steering vectors for control and developing latent space-based model calibrators for trust. Together, our contributions help demystify the latent spaces of language models and offer new insights into how to harness model internals to build more trustworthy language technology.
arXiv:2607.00009v1 Announce Type: new Abstract: Despite the remarkable proficiency of large language models (LLMs) in basic writing assistance, their utility in creative writing is fundamentally hindered by a persistent binary failure. This issue manifests as an oscillation between safe, surface-level editing, referred to as remedial polishing, and destructive, uncontrolled plot expansion. This dilemma defines a critical trade-off between narrative fidelity and descriptive intensity. We propose Loom, an assisted writing framework grounded in the narratological distinction between story and discourse. Loom employs a three-layer pipeline that operationalizes an intent-centered semiotic chain-of-thought to enforce precise control over narrative intent and rendering density. This architecture separates the generation of perceptual material from syntactic insertion, ensuring that enhancement occurs without violating the original event structure. Our comprehensive evaluation, which includes
arXiv:2607.00006v1 Announce Type: new Abstract: Beckmann & Butlin's (2026) ontological framework for the LLM individuation problem inherits an unargued cross-regime co-reference assumption from the persona-vectors literature: that the same direction picks out the same content under prompt-conditioning, gradient-descent fine-tuning, and inference-time steering. We present four empirical wedges from persona-topology experiments on Qwen3-4B-Instruct and Mistral-7B-Instruct-v0.2 - non-collinearity of prompt-extracted vectors and fine-tune basins; fictional personas displacing the model along real-anchor directions more strongly than real anchors do; contradictory-valenced mixtures biased toward a training-history-determined attractor; and asymmetric compositional algebra under inference-time arithmetic versus fine-tune-time chimera training - that jointly undermine the assumption. We propose regime-indexed individuation: the identity unit for representational content is a (vehicle, regime)
arXiv:2512.18303v2 Announce Type: replace-cross Abstract: Despite technical advancements, the human factor remains cybersecurity's most exploited vulnerability. Current research acknowledges this but remains fragmented, treating vulnerabilities as isolated, static traits. To address this, we introduce MORPHEUS, a holistic framework conceptualizing human-centric security as a dynamic, interconnected system. Grounded in the Cognition-Affect-Behavior (CAB) model and Attribution Theory, MORPHEUS consolidates 50 human factors influencing susceptibility to major cyberthreats (e.g., phishing, malware, password management, and misconfigurations). Beyond mere identification, the framework introduces a hierarchical Causal Pathway Architecture. Systematically mapping 302 empirical interactions (82.8% architecture-compliant), we reveal how cognitive, affective, and behavioral processes jointly shape security outcomes, distilling them into 12 recurring interaction mechanisms. MORPHEUS further links
arXiv:2410.22526v2 Announce Type: replace-cross Abstract: To effectively address potential harms from Artificial Intelligence (AI) systems, it is essential to identify and mitigate system-level hazards. Current analysis approaches focus on individual components of an AI system, like training data or models, in isolation, overlooking hazards from component interactions or how they are situated within a company's development process. To this end, we draw from the established field of system safety, which considers safety as an emergent property of the entire system. In this work, we translate System Theoretic Process Analysis (STPA) - a recognized system safety framework - for analyzing AI development and operation processes. We focus on systems that rely on machine learning algorithms and conduct STPA on three case studies involving linear regression, reinforcement learning, and transformer-based generative models. Our analysis explored how STPA's control and system-theoretic perspectiv
arXiv:2510.06452v3 Announce Type: replace Abstract: Recent advances in Large Language Models (LLMs) have introduced a new paradigm for software development, where source code is generated from natural language prompts. While this paradigm significantly boosts development productivity, building complex, real-world software systems remains challenging because natural language offers limited control over the code generation process. Inspired by the historical evolution of programming languages toward higher levels of abstraction, we advocate for a high-level abstraction language that gives developers greater control over LLM-assisted code writing. To this end, we propose Code Semantic Zooming (CodeZoom), a novel approach based on pseudocode that allows developers to iteratively explore, understand, and refine code across multiple layers of semantic abstraction. In a within-subjects user study (n=26), our method matches a state-of-the-art coding agent, Claude Code, on usability while produ
arXiv:2607.01034v1 Announce Type: cross Abstract: Large language model (LLM)-based conversational agents (CAs) are now ubiquitous, creating new opportunities for AI-mediated behavior change. Their capacity to project nuanced personalities and adopt diverse metaphorical roles raises a design question: how should an agent's persona and personality be calibrated to the moment? Recent evidence suggests that (i) moderate personality expression outperforms low or high extremes on trust, enjoyment, and intention to adopt in goal-oriented tasks, and (ii) context-appropriate metaphors outperform static one-note assistants on user experience and uptake. Yet most CAs still fix both persona and style, risking misalignment when dynamics, urgency, and formality vary, for example in medical information seeking, fitness coaching, and reflective learning. We propose a Fluid Personality Framework that jointly adapts (1) the agent's metaphorical persona, such as coach, tutor, librarian, or tool, and (2)
arXiv:2607.00968v1 Announce Type: cross Abstract: Emotion recognition in natural language is a foundational challenge in affective computing, with critical implications for human-computer interaction, mental health support, and conversational AI. This paper presents a rigorous, unified zero-shot evaluation of three leading commercial large language models: Claude (claude-sonnet-4-6), ChatGPT (GPT-5.4), and Gemini (gemini-2.5-flash). The models were queried through their respective production APIs as of April 2026 on a fine-grained 13-class emotion classification task. Using a stratified 1,000-sentence sample from the boltuix/emotions dataset, which comprises 131,306 sentences across 13 categories, a single uniform prompt with no exemplars was applied identically across all models. Gemini achieves the highest accuracy (39.9%) and macro-F1 score (0.363), followed by GPT-5.4 (38.8%, macro-F1 = 0.291) and Claude (38.0%, macro-F1 = 0.159). All models excel on sarcasm and desire while consis
arXiv:2607.00309v1 Announce Type: cross Abstract: We present a real-time musical interface that converts natural-language scene descriptions into evolving procedural soundscapes. A performer types a prompt such as "warm jazz cafe at midnight" and steers it through direct parameter adjustments - stepping brightness down, switching a rhythm style - each producing a predictable, audible shift without re-prompting. Where GPU-bound text-to-audio systems synthesize monolithic waveforms, our instrument generates human-readable configurations over a categorical schema, enabling fine-grained performer control; most valid combinations are designed to sound musically coherent. Three interchangeable backends - embedding retrieval for sub-second CPU-only use, hosted LLMs via API, and a fine-tuned 270M local model - all emit the same schema. A live generator architecture continuously emits audio while resolving new instructions in the background, crossfading seamlessly when ready; even when an LLM t
arXiv:2607.00211v1 Announce Type: cross Abstract: Epistemic thinking plays a central role in students' learning processes when applying generative artificial intelligence (GenAI), particularly in programming contexts where learners must construct queries, evaluate and validate AI-generated outputs, and regulate problem-solving strategies. This study introduces the conceptual framework of Epistemic AI Literacy (EAIL), reframing AI literacy as a process-oriented epistemic phenomenon that emerges through dynamic human-AI interactions across different domains. Drawing on the AIR (epistemic aims, ideals and reliable epistemic processes) framework, this study examines how epistemic aims and epistemic processes are enacted in GenAI-supported co-programming activities and explores scalable approaches for operationalizing these constructs in interaction data. Using a large dialogue dataset of human-AI co-programming, this study identifies observable dimensions of epistemic aims (i.e., mastery-o
arXiv:2607.01214v1 Announce Type: new Abstract: Statistical visualization is usually treated as a visual medium, but data can also be touched. Three dimensional printed tactile graphs let blind and low vision students feel distributions, trace trends, and explore relationships through direct haptic interaction. Yet classroom scale use remains limited because producing each graph in CAD software requires specialized skill and hours of manual work. We address this bottleneck as a software problem through a three layer reusable pipeline in about 1500 lines of JavaScript. The first layer derives tactile design parameters automatically from plate dimensions using tactile perception research. The second provides shared chart scaffolding and five modular builders for scatter, bar, histogram, line, and box plots. The optional third layer uses a multi-modal large language model to extract structured chart specifications from uploaded images, with mandatory teacher review before print generation
arXiv:2607.00989v1 Announce Type: new Abstract: Semantic trajectory analysis has recently emerged as an approach for modeling human movement by capturing implicit patterns and behaviors through semantic information (e.g., visitors' profiles and goals) beyond raw spatial paths to better understand why people move in certain ways. However, analyzing semantic trajectories in real-world scenarios remains challenging, as collecting high-quality data is costly and often lacks rich semantic information. Meanwhile, existing simulation tools require substantial technical expertise, which makes them difficult for practitioners to adopt. To address these limitations, the paper proposes ${SenseWalk}$, an interactive system that supports simulating semantic trajectories by LLM-powered agents. We develop a simulation workflow that combines LLMs and the social force model to balance physical plausibility and semantic coherence. A user-friendly interface is designed to facilitate users in customizing
arXiv:2607.00979v1 Announce Type: new Abstract: This study examined students' experiences with mixed-reality applications and physical toolkits in Engineering Mechanics to inform design guidelines for educational tools. In a user study with 24 participants, we compared classroom instruction alone, classroom instruction with a mixed-reality application, and classroom instruction with physical toolkits. Thematic analysis of participant feedback revealed that learners' workflows and engagement with fundamental mechanics problems varied across instructional modalities. Participants valued multimodal and interactive experiences that combined visualization with hands-on interaction, while reporting challenges with complex or unclear visualizations. These insights support the human-centered design of mixed-reality and physical tools for engineering education.
arXiv:2607.00969v1 Announce Type: new Abstract: Visual analytics (VA) plays an increasingly important role in supporting machine learning (ML) workflows. In the field of visualization, such approaches and techniques are referred to as VIS4ML. While ML models are mostly learned automatically, the corresponding ML workflows receive a variety of human inputs, such as data labelling, feature engineering, model architecture designing, hyper-parameter tuning, and so on. In this work, we surveyed over 200 VIS4ML papers to gain an understanding of how humans inject their knowledge into ML workflows through interactive visualization. We collected a corpus of VIS4ML papers from the IEEE VIS conferences in the past decade. We developed a coding scheme to facilitate the literature research from four perspectives: characteristics of ML, visualization, interaction, and actions. The analysis of the coded dataset allows us to observe different pathways that transfer human knowledge to ML workflows via
arXiv:2607.00542v1 Announce Type: new Abstract: The intersection of AI, healthcare, and visualization is evolving rapidly, posing challenges that cut across disciplinary boundaries and resist easy resolution. The Visual Analytics in Healthcare workshop (VAHC), co-located every other year at the IEEE VIS conference and the AMIA (American Medical Informatics Association) annual conference, has served as a forum to connect the visualization and medical informatics community since 2010. In 2025, to celebrate the 16th edition, we used the workshop as an opportunity to consolidate the community's collective experience (and expertise) and identify Grand Challenges where the field should prioritize going forward. We combined thematic coding of the 15 accepted VAHC workshop papers with structured group discussions among more than 40 participants, organized around three major themes: "Technical innovation vs. clinical reality", "Human-centered and scalable VAHC", and "From foundations to actiona
arXiv:2607.00533v1 Announce Type: new Abstract: As AI takes on more software work, the line between human and AI effort is shifting. Where developers draw that line around AI autonomy bears on how we design tools and roles that preserve meaningful work. Drawing on cognitive appraisal theory, work design, and automation research, we conducted a mixed-methods study of 448 professional developers at Microsoft to investigate their accepted levels of AI autonomy across software engineering work. Most developers accepted AI producing work under their oversight, although accepted autonomy varied substantively across tasks and individuals. Acceptance was lowest for identity-defining, human-facing, and design-oriented work, and higher among developers with more AI experience and risk tolerance. Task accountability was associated with lower odds of allowing AI to act on developers' behalf, whereas task identity was associated with lower odds of granting AI decision-making autonomy. Task demands
arXiv:2607.00518v1 Announce Type: new Abstract: Conventional Mixed Reality (MR) workspaces are frequently organized in cockpit-like layouts, where multiple floating windows surround the user. While this configuration facilitates access to digital content, it often induces occlusion, reducing understanding of the physical environment and limiting access to real-world objects. To overcome this challenge, we present the Contour-Adaptive Mixed Environment Overlays (CAMEO), a contour-adaptive MR interface that drapes virtual windows onto physical surfaces. This design integrates digital content with nearby items, thereby improving users' visual access to background objects and supporting interaction with them. We evaluate CAMEO in two controlled studies. The first demonstrates that draping reduces hand-movement detours relative to flat mid-air surfaces, enabling more direct interaction with nearby items. The second shows that controlled window deformation does not significantly impair text
arXiv:2607.00445v1 Announce Type: new Abstract: Proactive assistance with large language models (LLMs) has received growing attention in the human computer interaction (HCI) community. However, most past work on proactive LLMs' assistance has focused on adult users and task-oriented settings, leaving open how such systems could support children, whose interests and needs are often expressed through gaze and other nonverbal behaviors rather than explicit requests. In this study, we focus on two key challenges of proactive assistance in children's picture exploration: when to provide assistance and what assistance to provide based on children's nonverbal behaviors. To address these challenges, we introduce Ollie, a gaze-informed proactive artificial intelligence (AI) assistant that offers short narrative descriptions based on where a child is looking. Ollie uses children's gaze to estimate their attention, identify their current visual focus, and select a related picture region for the L
arXiv:2607.00420v1 Announce Type: new Abstract: Organizations increasingly deploy autonomous artificial intelligence (AI) systems for operational decisions, such as inventory replenishment. Yet fully granting override rights can degrade performance due to human bias and noise, while prohibiting them may overlook valuable private information. This raises a key question: How should override rights be structured to improve human supervision of autonomous AI? Methodology/results: We propose a constrained override policy that limits overrides per decision episode to enable selective filtering that prioritizes high-value overrides. We tested it through a randomized field experiment with 553 workers at a major Chinese smart vending machine retailer that manages more than 59,000 machines and 4,000 SKUs. Workers were assigned to no overrides, free overrides, or a two-per-machine limit on downward overrides. Free overrides reduce inventory by 1.95% but also cut sales by 1.19%. Constrained overri
arXiv:2607.00395v1 Announce Type: new Abstract: As generative AI is increasingly used by children and adolescents, there is a growing need for risk evaluation frameworks that account for child-specific harms. However, most existing safety evaluation frameworks focus on general user populations, often overlooking risks unique to younger users. To address this gap, we propose an evaluation framework that integrates expert-guided risk factors with real-world AI incident data for child safety. The framework identifies hazard categories from expert guidelines and AI incident databases and uses this information to construct a synthetic test set for model evaluation. Particularly, we apply the framework to the education domain and evaluate three Llama Guard models on their ability to detect unsafe user prompts. Our results show that current Llama Guard models struggle to identify education-related unsafe user prompts. We conclude by discussing how future work can extend the evaluation to addi
arXiv:2607.00239v1 Announce Type: new Abstract: In 3D visualizations of natural phenomena, improving aesthetics can provide measurable benefits, but often involves transformations that affect how the data is perceived. As a growing range of tools - including AI-based methods - make visual design and modification more accessible, it is increasingly important to understand trade offs and concerns when making these changes. We conducted an expert survey (N=95) with visualization researchers, practitioners, and domain scientists, investigating reactions to fifteen alterations spanning presentation-level adjustments (e.g., lighting, camera position) and data-level modifications (e.g., removing errors, filling gaps), applied by both humans and AI systems. Results show differences in perceived acceptability are driven by the transformation's meaning, regardless of whether it operates at the presentation or data level. Additionally, certain modifications were consistently judged as more permis
arXiv:2607.00103v1 Announce Type: new Abstract: Although textual framing in data visualizations is known to influence comprehension, recall, and perceptions of bias, its effects on viewers' emotional responses remain underexplored. Drawing on two widely studied framing strategies in political communication, we examine how episodic framing (foregrounding a specific event) versus thematic framing (foregrounding broader trends) affects emotional and attitudinal responses to visualizations. We conducted a preregistered, between-subjects online experiment (N = 800) in which participants viewed identical visualizations of U.S. mass shooting data that varied only in textual framing: a thematic title, a thematic title with annotation, or an episodic title paired with the same annotation. Results show that episodic framing elicited significantly more negative emotional valence than both thematic conditions. In contrast, adding an annotation to a thematic title did not alter emotional impact. Wh