Search research

Midway through an ordinary pretraining run, a small language model learns the pronoun-gender rule: cued with a girl's name ("Sue cried because"), it resolves the next pronoun to she, generalizing to held-out probes (0.94 by step 925). By step 3,500 the same model scores near zero on the same probes, although the rule's evidence is still in the training data. We call this within-run reversal natural ungrokking: the corpus decides, with no trace in the loss curve, which learned rules a model keeps. Which rules survive is predictable from one corpus statistic: how often the training stream shows the rule winning. Across un-intervened runs (two corpora, three budgets, three seeds), support frequency decides a rule's fate; the data-to-parameter ratio only modulates how deeply a doomed rule falls. The same emerge-then-collapse dynamics appear in public Pythia checkpoints, collapse depth ordered by model scale as predicted. The forgetting is a displacement: a competing surface pattern out-competes the rule, and the log-probability margin between them crosses zero within 100 training steps of the behavioral collapse. Control over this fate is asymmetric: the same edit that destroys a rule on demand cannot restore it. Flipping support to counter-evidence in place kills the rule with monotone dose-response in two unrelated rules; but injecting support back, even to 450 times the level that naturally sustains it, buys no recovery. Every confirmatory threshold and prediction was pre-registered before the data it governed was read.

Diffusion Transformers (DiT) trained with flow matching in a VAE latent space have unified visual generation across images and videos. A natural next step toward a single architecture for both generation (visual synthesis) and understanding (text generation) is to apply this framework to language modeling. We propose TextLDM, which transfers the visual latent diffusion recipe to text generation with minimal architectural modification. A Transformer-based VAE maps discrete tokens to continuous latents, enhanced by Representation Alignment (REPA) with a frozen pretrained language model to produce representations effective for conditional denoising. A standard DiT then performs flow matching in this latent space, identical in architecture to its visual counterpart. The central challenge we address is obtaining high-quality continuous text representations: we find that reconstruction fidelity alone is insufficient, and that aligning latent features with a pretrained language model via REPA is critical for downstream generation quality. Trained from scratch on OpenWebText2, TextLDM substantially outperforms prior diffusion language models and matches GPT-2 under the same settings. Our results establish that the visual DiT recipe transfers effectively to language, taking a concrete step toward unified diffusion architectures for multimodal generation and understanding.

Automatic scientific discovery has long been a goal of computational scholars - a machine that can discover nature's secrets on its own, moving computational systems beyond data-fitting tools toward the generation and refinement of mechanistic models of the universe. Recent advances in symbolic regression (SR) and large-language-model (LLM)-based agents suggest that such systems can recover equations from data, incorporate domain priors, and automate parts of the research workflow. However, most existing approaches either focus on narrow equation-discovery benchmarks or broad end-to-end automation pipelines, while biological systems remain comparatively underexplored. Here, we introduce the MEDA system, an LLM- and SR-powered agentic framework for discovering ordinary-differential-equation (ODE) models of biological and biologically inspired dynamical systems. MEDA retrieves background knowledge, defines admissible variables, generates mechanistic constraints, proposes candidate ODEs, and fits and evaluates them. We evaluate it across canonical model retrieval, reasoning-based extrapolation to unseen variants, and open-ended discovery, with and without experimental data. Across these settings, MEDA recovered the correct state variables, achieved strong structural recovery in retrieval and extrapolation tasks, and produced biologically plausible discovery-oriented models. Ablation and robustness analyses show that knowledge-guided formalization and mechanistic constraints are load-bearing components, whereas numerical fitting alone can preserve trajectory-compatible but biologically incorrect equations.

Slang is a central component of everyday language, reflecting linguistic creativity, social identity, and cultural change, yet its dy- namic and non-standard nature makes it difficult to model computationally. We present the first large-scale computational study of slang.gr, a crowdsourced lexicon of Greek non-standard language, combining lexical content, user-generated tags, and interaction data. To enable the systematic analysis, we map noisy folksonomic tags to a structured multi-layer taxonomy capturing both semantic categories and sociolinguistic metadata. Using this representation, we analyze the linguistic structure of Greek slang and the behavior of its contributor community. We find that slang is strongly centered on person-related and evaluative language, exhibits high morphological creativity, and is shaped by highly skewed participation with short user lifespans and overlapping communities. Building on these signals, we introduce a community-based confidence score for definitions that integrates user roles, interaction patterns, and moderation signals. Our results show that taxonomy-based representations improve interpretability while retaining meaningful aspects of behavioral structure, enabling a more structured and interpretable analysis of confidence signals. Overall, this work establishes slang.gr as a computational resource for non-standard Greek and provides a foundation for sociolinguistic NLP, bias analysis, and the study of informal language in LLMs.

Standard RLHF relies on transitive scalar rewards, failing to capture the cyclic nature of human preferences. While some approaches like the General Preference Model (GPM) address this, we identify a theoretical limitation: their implicit formulation entangles hierarchy with cyclicity, failing to guarantee dominant solutions. To address this, we propose the Hybrid Reward-Cyclic (HRC) model, which utilizes game-theoretic decomposition to explicitly disentangle preferences into orthogonal transitive (scalar) and cyclic (vector) components. Complementing this, we introduce Dynamic Self-Play Preference Optimization (DSPPO), which treats alignment as a time-varying game to progressively guide the policy toward the Nash equilibrium. Synthetic data experiments further validate HRC's structural superiority in mixed transitive--cyclic settings, where HRC converges faster and achieves higher accuracy than GPM. Experiments on RewardBench 2 demonstrate that HRC consistently improves over both BT and GPM baselines (e.g., +1.23% on Gemma-2B-it). In particular, its superior performance in the Ties domain empirically validates the model's robustness in handling complex, non-strict preferences. Extensive downstream evaluations on AlpacaEval 2.0, Arena-Hard-v0.1, and MT-Bench confirm the efficacy of our framework. Notably, when using Gemma-2B-it as the base preference model, HRC+DSPPO achieves a peak length-controlled win-rate of 44.75% on AlpacaEval 2.0 and 46.8% on Arena-Hard-v0.1, significantly outperforming SPPO baselines trained with BT or GPM. Our code is publicly available at https://github.com/lab-klc/Hybrid-Reward-Cyclic.

Vision-language alignment powers open-vocabulary recognition, retrieval, and LVLM grounding, yet natural captions are often underspecified, making similarity brittle and overly confident under paraphrase and omitted details. We aim to learn representations whose matching is stable across caption views and whose confidence reflects how strongly text constrains an image. We propose Text as Partial Constraint (TPC), a core-residual alignment framework that treats multi-view captions as incomplete supervision. It distills a consensus semantic core as the alignment target, learns a single-view core predictor for standard inference with one query, and explicitly discourages vision-language similarity from depending on the orthogonal unsaid residual. An uncertainty-aware contrastive objective further softens alignment when caption views disagree, reducing overconfident updates under weak language constraints. Across zero-shot recognition and adversarial robustness, TPC achieves 81.42/64.05 Top-1 clean/robust accuracy on ImageNet and 76.19/52.03 on an Avg-14 transfer suite, while improving LVLM transfer with 85.16 POPE F1 and 59.57 OKVQA accuracy under an LLaVA-1.5-7B stack. These results suggest that modeling text as a partial constraint is a practical and principled route to more reliable vision-language representations under underspecified language supervision.

Existing circuit discovery methods rely on templated tasks with clean counterfactuals, limiting their use on diverse natural text. We adapt Contextual Decomposition for Transformers (CD-T) for unstructured settings via label-balanced activation means and task-directional relevance scoring, enabling counterfactual-free circuit discovery. We leverage the discovered circuits for Circuit-Targeted Supervised Fine-Tuning (CT-SFT), restricting parameter updates to task-relevant heads and LayerNorm. Experiments on NusaX cross-lingual sentiment transfer show that CT-SFT is highly competitive for low-resource adaptation. While non-circuit sparse updates and full fine-tuning sometimes match target accuracy through capacity recruitment, CT-SFT most consistently avoids catastrophic forgetting, preserving source-language and related-task performance. Extensions to XNLI support the source-retention and intervention findings on a harder task and two model families, showing that circuit-targeted adaptation provides a more controlled, intervention-supported alternative to global fine-tuning.

Protein language models (pLMs) can generate novel protein sequences with properties beyond those observed in nature, yet the mechanisms underlying protein generation remain poorly understood. Existing mechanistic interpretability methods based on sparse autoencoders and transcoders primarily focus on protein representation learning models and do not capture the computation required for autoregressive generation. Here, we introduce ProGenMech, a mechanistic interpretability framework for generative protein language models that extends cross-layer transcoders (CLTs) to ProGen3, a sparse Mixture-of-Experts model trained for both causal generation and span infilling. Unlike per-layer approaches, CLTs reconstruct each layer using sparse latent variables from all preceding layers, enabling faithful recovery of inter-layer generative computation. We further develop a zero-shot circuit discovery framework to identify sparse latent circuits responsible for protein generation and fitness prediction. In causal generation and zero-shot fitness estimation tasks, ProGenMech outperforms local transcoder baselines in recovering ProGen3's probability distribution and functional scoring behavior, while matching the original model's generative distribution in span infilling tasks. Moreover, the recovered circuits reveal biologically meaningful motifs and functional regions associated with conserved sequence patterns and protein fitness landscapes, establishing a foundation for interpretable and steerable protein generation.

Text-to-speech (TTS) evaluation for low-resource non-Latin-script languages can fail when it relies on a single ASR round-trip word error rate (WER). A system may produce no audio, speak a neighbouring language, preserve target script text only in an ASR transcript, or sound unnatural to native listeners. We introduce INSV (Intelligibility, Naturalness, Script fidelity, and Verification), a reporting framework that separates these cases. This paper reports INSV-A, the automated screening subset: synthesis completion, ASR WER/CER, transcript Script Fidelity Rate, and audio language identification. Native MOS and phonetic annotation are specified but not claimed in this release. We instantiate INSV-A as PashtoTTS-Bench, a dated benchmark for Pashto TTS. The April-May 2026 run evaluates Edge GulNawaz, Edge Latifa, OmniVoice clone, OmniVoice auto, and an Urdu negative control on 200 FLEURS and 200 filtered Common Voice 24 prompts. Under the independent omniASR_CTC_300M_v2, OmniVoice auto has the lowest WER (24.1% FLEURS, 27.4% CV24), followed by Edge GulNawaz (32.8%, 39.5%), Edge Latifa (35.6%, 47.7%), and OmniVoice clone (45.4%, 34.8%). WER below the natural-speech baseline reflects clean synthetic audio and should not be read as better than native speech. Whisper Large V3 returns 0.0% Pashto labels on checked Pashto TTS audio, while MMS-LID-4017 and SpeechBrain VoxLingua107 separate Pashto outputs from the Urdu control. The release provides provider metadata, per-sentence scores, LID audits, failure logs, and scripts for adding systems.

Hyperbolic vision-language models (VLMs) represent image and text features in a geometry naturally suited to hierarchy, but their adaptation to downstream tasks has largely relied on fixed prompts. Existing prompt learning methods, meanwhile, treat class labels as a flat set and do not exploit available taxonomic structure. We address this gap with a hierarchical prompt learning plug-in for frozen hyperbolic VLMs. Given a fixed offline parent-class hierarchy, it augments a class prompt learner with a separate parent prompt learner, parent-level supervision, hyperbolic entailment regularization, and parent-feedback logit fusion. We instantiate the method with CoOp, CoCoOp and MaPLe, yielding HyPLO, CoHyPLO and MaHyPLO. Across the standard 11-dataset benchmark, all variants improve base-to-new generalization and cross-dataset transfer, and remain comparable to their prompt learning baselines under domain shift. Six hierarchical metrics and embedding analyses show that the method produces more taxonomically consistent predictions and induces a hierarchy-consistent organization of parent, class, and image embeddings in hyperbolic space. Its gains are largest when novel classes must be placed within a fixed taxonomy, and smallest for fine-grained confusions among sibling classes or shifts affecting only the image distribution.

Large Language Models (LLMs) are frequently confident, eloquent, and well versed. A natural question arises: do they know what they don't know? To answer this question, we borrow the concept of epistemic honesty and develop a novel metric to systematically evaluate whether an LLM appropriately acknowledges the boundaries of its knowledge. In this work, we introduce the Epistemic Honesty Quotient (EHQ), which reports three observable sub-scores across two operational axes (epistemic restraint and substantive-answer calibration), and construct EHQ-3000, a 3,000-question benchmark spanning Fabricated Entity, Post-Cutoff Event, Hyper-Niche True, and Context-Conditioned Questions. From a frozen registry of 21 model API routes, 15 completed the protocol after endpoint and eligibility checks; 14 entered the confirmatory analysis because severe provider-side truncation made one route's score indeterminate. The study reveals substantial variation across models, including a difference that can not be explained by their capability to extract explicitly available information. Composite EHQ ranges from 0.31 to 0.81 across the analysed panel, despite near-ceiling performance on the document-grounded capability probe. The two restraint criteria overlap strongly under the present category composition, whereas substantive-answer calibration varies across models and does not reliably co-vary with restraint; however, the small panel leaves substantial uncertainty. Thus, EHQ reveals behavioral differences that are not visible to conventional correctness-based assessment, while also showing why dataset composition, provider behavior, and confidence elicitation must remain part of the interpretation.

Films communicate through deliberate creative choices, including lighting, color, composition, editing, dialogue, music, and sound. Humans naturally interpret these signals as directorial intent, yet current multimodal large language models (MLLMs) are evaluated almost exclusively on understanding what happens rather than why it is presented that way. We introduce TAKE 85, the first benchmark for directorial-intent understanding, comprising 398 short films (85 hours) with expert-verified question-answer pairs spanning global and fine-grained visual and audio intent. Through controlled modality ablations, TAKE 85 enables systematic evaluation of multimodal reasoning. Experiments on state-of-the-art MLLMs reveal a substantial gap between perceptual recognition and intentional understanding: while models accurately describe events and narratives, they consistently fail to infer the communicative role of filmmaking decisions. Our results establish directorial intent as a previously overlooked dimension of multimodal understanding: even the strongest model reaches only 58 out of 100, and our ablations show that no input modality is sufficient on its own. All code, Q&As, and models are publicly available from https://github.com/KaiShinozakiConefrey/Take-85

LLM agents increasingly modify their own prompts, tools, middleware, resources, and execution harnesses at runtime. Such self-evolution can improve capability, but a successful mutation may leave persistent effects that cannot be safely reversed in states different from the one in which it was created. We introduce EvoUndo, a framework for representing, synthesizing, diagnosing, and independently verifying recoverability of model-generated self-modifications across counterfactual states. Across 600 unseen one-shot self-evolution tasks, we identify 197 capability-improving mutations that fail recoverability verification. Under the original recovery representation, conventional repair strategies recover 0/197 of these natural failures. Deterministic oracle analysis recovers 48/197 under the original recovery language L0, while the extended recovery calculus increases empirical oracle recovery to 191/197. A protocol-locked 2x2 grounding-by-expressivity intervention then separates two bottlenecks: exact state-address grounding increases successful recovery from 0/48 to 38/48 (79.2%) when the original language is sufficient, while extending the recovery language enables recovery on 142/143 (99.3%) failures in the oracle-defined S1 stratum. On the primary gpt-oss-120b backbone, adding exact-address diagnostics to the richer language reduces recovery to 133/143 (93.0%); a Qwen3.8-27B replication preserves the grounding and expressivity effects but not this negative interaction, indicating that the latter is model-dependent. These results indicate that reliable agent self-evolution requires co-designing verification, state grounding, witness semantics, and recovery-language expressivity rather than relying on iterative prompting alone.

Whole slide image (WSI) analysis is central to computational pathology, with multiple instance learning (MIL) emerging as the standard pipeline for slide-level diagnosis. However, conventional approaches formulate WSI diagnosis as a flat classification task over discrete labels, contradicting the inherently hierarchical, coarse-to-fine nature of clinical reasoning. Although recent hierarchical classifiers and vision-language models (VLMs) have sought to address this structural gap, they either fail to capture semantic continuity between related diagnoses or suffer from unconstrained text generation that produces taxonomic hallucinations and parent-child label violations. To address these limitations, we propose TaxoMIL, a taxonomy-constrained framework that reformulates WSI diagnosis as a multi-granularity text generation task. TaxoMIL utilizes a dual-head Transformer decoder to generate coarse- and fine-level diagnostic text, and introduces taxonomy-guided objectives that explicitly structure the label embedding space and strictly ground slide-level visual representations within the clinical taxonomy. Extensive experiments across three diverse WSI datasets demonstrate that TaxoMIL consistently outperforms state-of-the-art MIL classifiers and VLM-based generative methods, yielding accurate and hierarchy-aware diagnostic predictions. The code is released at https://github.com/QuIIL/TaxoMIL

Large Language Models (LLMs) are increasingly embedded in social robots to support natural group interactions, yet their role in complex multi-party settings remains underexplored. In particular, it is unclear how LLM-driven robots decide when and why to intervene in group conversations. This paper investigates the intervention explanations generated by an LLM-based orchestrator in a multi-party interaction involving three human participants and two robots. We conducted a between-subjects study with 24 groups (66 university students), comparing a homogeneous condition (two robots with the same role, i.e., a mover) and a heterogeneous condition (two robots with different roles, i.e., a mover and an opposer). At each conversational turn, the LLM orchestrator decided whether to intervene and generated a textual explanation of its decision. We performed a thematic analysis of 610 intervention explanations, identifying five recurring themes. Results show that explanations are facilitation-oriented, emphasizing agreement, participation, and interaction flow. While patterns remain stable across conditions, role differentiation emerges: the mover supports coordination, whereas the opposer drives goal-oriented interventions. These findings contribute to explainable AI by characterizing how LLM-driven systems justify intervention decisions in real-time, multi-party human-robot interaction.

Does intelligence require the ability to reason about phenomena beyond direct experience? It is natural to suspect that some complex thought cannot be captured through language alone. However, of particular concern to this work, is whether visualising counterfactual events can complement language as a mechanism for complex thought. We ask whether LLMs can be trained to utilise such visualisation mechanisms, in a way that benefits their reasoning abilities. Motivated by this question, we propose Einstein World Models. EWMs are a blueprint for LLM-based reasoning systems that place visual-temporal rollouts inside the reasoning trace, allowing them to reason in ways that text alone may not support well. In an EWM, the LLM calls a world-module (not to be confused with a world model), to produce short rollouts of scenes under consideration. The returned rollout is treated not as the answer, but as an inspectable hypothesis that can support later reasoning. Einstein World Models extend the capability of LLMs for tool calling (such as web search or code execution), into the domain of visual thought experiments.

OpenClaw has rapidly emerged as a transformative artificial intelligence (AI) agent framework, and its ability to autonomously execute complex, multi-step tasks has attracted an ever-growing and diverse user base. However, this capability comes with significant risks. While existing research has made important strides in characterizing these threats, such work is predominantly directed at technically sophisticated audiences. It remains largely inaccessible to non-technical users. This demographic now makes up an increasingly large and underserved portion of the community, yet it is these very users who most urgently need practical and straightforward guidance. In response, we bridge this gap through a series of interconnected efforts designed to lower the risk barrier for non-technical OpenClaw users. First, we identify and categorize seven core risks that OpenClaw users may encounter in daily usage, explaining each in plain language so that non-technical users can readily grasp the nature and potential consequences of these threats. Second, for each identified risk, we distill a set of corresponding defensive strategies into clear and actionable operational steps that are easy to follow. Third, to make protection even easier, we provide a companion OpenClaw Skill that automates key security configurations, enabling users to safeguard their systems with minimal manual intervention. Through this work, we demonstrate that safeguarding against the risks of intelligent agents need not be the exclusive domain of security experts, and that non-technical users can meaningfully participate in reducing these risks through simple, practical actions.

Autoregressive (AR) image generation models are highly expressive but computationally intensive, motivating effective model compression. Knowledge distillation (KD) is a natural approach for model compression and has been widely studied in language modeling, yet its behavior in visual AR generation remains underexplored. In this work, we present the first systematic study of distillation strategies for AR image models. Our analysis shows that while standard distillation can yield meaningful gains, recent methods developed for language do not directly transfer to images: long decoding horizons and visual token ambiguity make teacher supervision unreliable especially under student-conditioned contexts. To address this, we propose VarKD, a distillation framework for visual autoregressive models that distills on student samples while selectively applying teacher supervision and reducing token-level ambiguity. Experiments on ImageNet across multiple AR backbones show that VarKD consistently outperforms prior distillation baselines, narrowing the gap to large-scale models.

Human language comprehension unfolds sequentially: each word is processed in the context of those that came before, and the interpretation builds incrementally over time. Surprisal, the negative log probability of a word given its context, has been the dominant predictor of incremental processing cost. But surprisal reduces rich sequential representations to a single scalar at each word, discarding information about the direction in which the interpretation has been evolving. Dynamical-systems approaches suggest that the trajectory of the evolving interpretive state, not just its position at each moment,should shape processing, and language itself may have local momentum, since speakers plan utterances a few words at a time. We introduce trajectory extrapolation error: at each word, we fit a linear trajectory to the preceding hidden states of a transformer language model and measure deviation from the extrapolated path. On the Natural Stories corpus, this measure is nearly orthogonal to surprisal (r = .044) and independently predicts self-paced reading times. The effect is especially pronounced in garden-path sentences, strengthens with model scale (GPT-2 Small to Large), and replicates across architectures with different positional encoding schemes (GPT-2 vs. Pythia/RoPE). A displacement control shows the effect is not reducible to representational change magnitude: displacement and extrapolation error predict in opposite directions. These findings reveal two dissociable components of processing cost: word-level prediction error (surprisal) and sensitivity to the local momentum of the unfolding interpretation (trajectory extrapolation error).

In language reasoning, longer chains of thought consistently yield better performance, which naturally suggests that visual latent reasoning may likewise benefit from longer latent sequences. However, we discover a counterintuitive phenomenon: the performance of existing latent visual reasoning methods systematically degrades as the latent sequence grows longer. We reveal the root cause: Information Gain Collapse -- autoregressive generation makes each step highly dependent on prior outputs, so subsequent tokens can barely introduce new information. We further identify that heavily pooled ($\geq 128\times$) image embeddings used as supervision targets provide no more signal than meaningless placeholders. Motivated by these insights, we propose SCOLAR (Self-COnsistent LAtent Reasoning), which introduces a lightweight detransformer that leverages the LLM's full-sequence hidden states to generate auxiliary visual tokens in a single shot, with each token independently anchored to the original visual space. Combined with three-stage SFT and ALPO reinforcement learning, SCOLAR extends acceptable latent CoT length by over $30\times$, achieves state-of-the-art among open-source models on real-world reasoning benchmarks (+14.12% over backbone), and demonstrates strong out-of-distribution generalization.

Remote sensing vision-language models (RS-VLMs) face a fundamental mismatch with natural-image counterparts: the same geographic object exhibits radically different visual evidence across ground sampling distances (GSDs) spanning multiple orders of magnitude. Yet existing RS-VLMs often discard GSD or inject it as a discrete text token, forcing a single static parameter set to absorb the entire scale spectrum. We introduce ScaleEarth, a parameter-efficient fine-tuning framework built on Qwen3-VL that treats GSD as a continuous conditioning variable governing the model's computation path. At its core, CS-HLoRA (Continuous Scale-Conditioned Hyper-LoRA) modulates the LoRA low-rank subspace through a GSD-driven gate, enabling the model to dynamically route computation by physical scale. To remove reliance on sensor metadata at deployment, we pair CS-HLoRA with SSE-U, a lightweight heteroscedastic sub-head that predicts GSD and its uncertainty from visual features. To provide matching supervision, we construct GeoScale-VQA, a 1.5M-sample scale-layered RS-VQA corpus whose question-answer generation is conditioned on the same physical scalar that drives CS-HLoRA, forming a closed method-data loop. Trained with QLoRA on an 8B backbone, ScaleEarth achieves state-of-the-art results on remote-sensing benchmarks covering diverse Earth-system tasks, including XLRS-Bench and OmniEarth-Bench.

Evaluating the writing capabilities of large language models (LLMs) remains a significant challenge due to the multidimensional nature of writing skills and the limitations of existing metrics. LLM's performance in thousand-words level and open-ended writing is inadequately assessed by traditional reference-based metrics or modern LLM-as-a-judge methods. We propose Tree-of-Writing (ToW), to resolve the implicit inconsistency often found when LLM-as-a-judge aggregates all sub-features in text evaluation. ToW incorporates a tree-structured workflow by explicitly modeling the aggregation weights of sub-features. We also present HowToBench, a large-scale Chinese writing benchmark encompassing 12 genres and 1302 instructions across three task categories: contextual completion, outline-guided writing, and open-ended generation. ToW successfully mitigates the biases, achieving a 0.93 Pearson correlation with human judgments. Furthermore, we detect that both overlap-based text generation metrics and popular LLM-as-a-judge practices are vulnerable to textual disturbances, while ToW is robust to them. We also uncover a negative correlation between input length and content-related scores in the Guide task, showcasing that it cannot be simply improved by input-side information piling.

Vision-language models (VLMs) have demonstrated strong performance in visual question answering with natural images. However, they continue to struggle with scientific diagrams, which are designed to convey functional or relational meaning rather than literal scenes. We therefore introduce a framework for generating large-scale diagram-grounded instruction data by leveraging terminology derived from scientific curricula. Our approach systematically extracts domain concepts, synthesizes atomic facts, retrieves relevant diagrams from the web, and generates multimodal supervision in the form of diagram captions and multiple-choice questions. Using this pipeline, we construct SciGram, a dataset of over 194K diagrams and 1.4M visual instructions across life, earth, and physical sciences. Despite relying on noisy web data and synthetic annotations, models fine-tuned on SciGram achieve substantial improvements on diagram-centric benchmarks, including TQA, ScienceQA, and AI2D, outperforming or matching state-of-the-art VLMs while using fewer training instances. Furthermore, augmenting existing models such as LLaVA OneVision with SciGram establishes new state-of-the-art performance on diagram question answering. Our results highlight the effectiveness of terminology-grounded instruction generation as a general strategy for improving vision-language reasoning in scientific domains. To support future research in scientific diagram understanding, we release both the SciGram dataset and models.

Large Language Models are increasingly deployed in emotional-support contexts and crisis-related situations. Nevertheless, their cross-lingual abilities in these circumstances remain underexplored. Existing benchmarks emphasize multilingual performance but rarely examine crisis-related empathy and cultural grounding in low-to-mid-resource languages. We introduce SPLIT, a 500-prompt benchmark designed to evaluate LLM consistency in generating emotionally grounded responses across five categories: Stress, Panic, Loneliness, Internal Displacement, and Tension. We evaluate three technically diverse LLMs across three dimensions: Empathetic Accuracy, Linguistic Naturalness, and Contextual & Cultural Grounding. The framework aims to assess and compare the quality of LLM responses in both English and Ukrainian languages, as well as to explore the reliability of the LLM-as-a-jury paradigm. Our findings reveal that Gemini-2.5-Flash and LLaMA-3.3-70B-Instruct degrade when transitioning to Ukrainian, while DeepSeek-V3 remains comparatively stable within our benchmark. We additionally find that human and AI evaluators agree weakly on empathy and naturalness but diverge on cultural grounding. We further argue that producing Ukrainian text is not equivalent to producing Ukrainian emotional support. Our findings may assist in the future development of more culturally tailored benchmark designs, as well as encourage a stronger emphasis on human-centered evaluation.

The thinking-while-speaking paradigm aims to make AI communication more human. A key challenge is maintaining fluent speech while performing deep reasoning. Our method, InterRS, tackles this by inserting reasoning steps only during natural speech generation. This requires high-quality data where reasoning and speech are precisely aligned, and the length ratio are under controlled. We introduce a novel pipeline to generate such seamlessly interleaved audio data. To train our model, we combine interleaved SFT with refined data and reinforcement learning with two new rewards: a TA-Balance Reward to manage timing and thinking-answer ratio, and a Linguistic Quality Reward to refine expression. Experiments show our approach achieves 13% better performance on mathmatical and logic benchmarks while generating instant response like a spoken-language instruct model which outputs fast CoT response. Furthermore, our method generates more natural and fluent answers than prior methods.

Large language models increasingly rely on external tools to access up-to-date information, perform computation, and interact with the outside world. For autoregressive models, tool use naturally fits the generation process: the model emits a tool call, waits for the result, and then continues generating. Diffusion language models (dLLMs), however, reason by repeatedly refining many parts of their output in parallel, making this stop-and-resume interaction pattern unnecessarily restrictive. It can force tool decisions before the model's reasoning has stabilized, delay useful observations until a discrete call finishes, and introduce redundant refinement and tool execution, potentially hurting both task accuracy and inference efficiency. We introduce Continuous Interaction Diffusion (CID), a diffusion-native model--runtime architecture that integrates tool interaction into iterative denoising. CID separates a model-read-only fact channel, a thought channel represented by a Typed Cognitive Tensor, and a display channel. Information needs can emerge before a textual or JSON call is fully serialized, allowing perceptual bindings to launch external reads while denoising continues. Returned results are projected into the evolving thought state and can revise earlier cognition and display regions. Persistent bindings reuse static results without repeated external execution and refresh changing sources when needed. CID is designed to expose evidence earlier, overlap tool latency with model computation, reduce duplicate external work, and preserve useful computation after new evidence arrives. We formalize the architecture, runtime, and training objectives, and define an evaluation protocol for task quality and end-to-end efficiency. This first paper focuses on read-only tools and makes no empirical performance claims.

Modern manufacturing imposes six coupled demands on adaptive control: local decisions with global consequences, partial observability, nonstationarity, reflex speed response with long horizon effects, delayed and diffuse outcomes, and dynamics that resist explicit modeling. Cooperative multiagent reinforcement learning (MARL), posed as a Dec-POMDP under centralized training with decentralized execution, is a particularly natural formalism for these demands. This paper adopts a MARL centered scope and asks where large language models (LLMs) should augment, interface with, train, or, in the strongest competitive case, replace that coordination core. A taxonomy organizes the literature through four LLM attachment points: policy, reward design, communication between agents, and hierarchical planning. A conditional capability profile separates native mechanism, reported performance, formal guarantee, and engineering maturity, and a deployment readiness analysis identifies the evidence behind each role. These stages yield the principal contribution: a three layer MARL centered reference architecture, grounded in evidence, for semantic reasoning, adaptive cooperative control, and independently assured execution. The LLM-Augmented Dec-POMDP is a descriptive comparative notation for that architecture, recording four attachment choices without introducing a new decision process class or algorithm. Under the reviewed evidence, conventional MARL is better suited to frequent, structured, decentralized coordination after task specific training, whereas LLM components are promising for semantic interpretation, reward drafting, human interaction, and slower supervisory planning. Current LLM only manufacturing controllers do not yet establish equivalence for strict real time, decentralized, safety critical control; this conclusion is bounded by the available evidence and does not assert impossibility.

The quest for intuitive and natural human-robot interaction (HRI) remains a significant challenge in robotics. Traditional methods often rely on rigid, pre-programmed commands that limit the robot's expressiveness and adaptability. This paper introduces a novel framework that leverages the reasoning capabilities of Large Language Models (LLMs) to synthesize complex robotic actions from a rich tapestry of multimodal human inputs: natural speech, hand gestures, and music/sound beats. Our system architecture integrates a speech transcription model, a gesture recognition module, and a signal processing pipeline for beat detection. These processed inputs are contextualized using prompt templates and fed into a LLM. The LLM, informed by a predefined robot action space, reasons over the combined inputs to generate a coherent sequence of actions. This sequence is dispatched to an action queue for execution on a quadruped robot over ROS. The framework has ability to interpret and fuse semantic commands from speech, deictic information from gestures, and rhythmic cues from music. This work represents a step towards creating robots that can interact with humans in a more fluid, creative, and context-aware manner.

Most state-of-the-art sign language models are trained on interpreter or isolated vocabulary data, which overlooks the variability that characterizes natural dialogue. However, human communication dynamically adapts to contexts and interlocutors through spatiotemporal changes and articulation style. This specifically manifests itself in educational settings, where novel vocabularies are used by teachers, and students. To address this gap, we collect a motion capture dataset of American Sign Language (ASL) STEM (Science, Technology, Engineering, and Mathematics) dialogue that enables quantitative comparison between dyadic interactive signing, solo signed lecture, and interpreted articles. Using continuous kinematic features, we disentangle dialogue-specific entrainment from individual effort reduction and show spatiotemporal changes across repeated mentions of STEM terms. On average, dialogue signs are 24.6%-44.6% shorter in duration than the isolated signs, and show significant reductions absent in monologue contexts. Finally, we evaluate sign embedding models on their ability to recognize STEM signs and approximate how entrained the participants become over time. Our study bridges linguistic analysis and computational modeling to understand how pragmatics shape sign articulation and its representation in sign language technologies.

Scene text spotting remains challenging for arbitrarily shaped text instances such as curved signs and dense multi-oriented characters in natural images, where tightly coupled architectures propagate localization errors directly into recognition failures. We present a two-stage pipeline that combines multi-encoder segmentation with vision-language model recognition to address this problem. In the detection stage, six frozen vision encoders (CLIP, DINOv2, SigLIP, EVA-CLIP, SAM, and ConvNeXt) extract complementary features spanning semantic, spatial, and texture spectra, which are fused through a trainable hierarchical Feature Pyramid Network with channel attention and decoded via a deep-supervision Progressive Scale Expansion network to generate precise instance-level text masks. By keeping the encoders frozen, their independently learned feature spaces remain orthogonal during fusion, preventing the feature homogenization that degrades boundary precision in single-backbone detectors. The detection stage produces tight polygon masks that conform to the actual shape of curved and arbitrarily oriented text, rather than axis-aligned rectangles that inevitably include background content. In the recognition stage, these polygon-masked crops isolate the target text from surrounding clutter, allowing a Qwen3-VL-8B-Instruct model, fine-tuned via Low-Rank Adaptation on polygon-cropped scene text, to focus purely on reading the text without interference from neighbouring words or background noise. Without any synthetic pretraining data, our method achieves 91.99% detection F-measure and 85.86% end-to-end H-mean on CTW1500, setting a new state of the art and achieving strong performance on Total-Text and ICDAR 2015 without any synthetic training data. Code is available at https://github.com/doubleblind-afk/MEVL-STP