LLM 理论进展

3D Gaussian language fields provide an explicit, spatially grounded representation for 3D visual question answering (VQA), but their dense semantic features can require tens of thousands of embeddings per scene, resulting in substantial storage, memory, and inference costs. We investigate how much of this representation is actually necessary for downstream reasoning. Starting from a full embedding representation, we systematically sparsify its semantic embeddings, including the previously underexplored regime below a single image-equivalent block down to 8 visual tokens. We compare random, geometric, semantic, and joint spatial-semantic selection strategies and introduce an object-based sparsification method that distributes the token budget across detected object instances while retaining background context. Experiments on ScanQA and MV-ScanQA reveal substantial redundancy in dense Gaussian language fields. Strong VQA performance is retained with only a few hundred semantic embeddings, corresponding to less than 1% of the original representation. Object-based selection performs well relative to others, with only modest observed changes down to 256 tokens. At this budget, SparseTalk retains 0.80% of SplatTalk's 32,076-token inference input and 0.332% of the mean 77,207-Gaussian dense field, increasing inference throughput while reducing decoded-feature memory 125-fold.

Powerful technologies labeled ``AI''-without sufficient epistemic caution-are already reshaping political and private life, bringing both new dangers and new opportunities for citizen participation. These range from electoral and legislative engagement to the most ambitious form: political co-creation through citizens' assemblies. Large Language Models (LLMs) could support such processes through moderation, translation, facilitation, summarization, and writing assistance. But this potential remains largely unrealized. The Democratic Commons project takes a fundamentally interdisciplinary approach-from philosophy to computer science-to evaluate LLMs against five proposed democratic principles. At its core, the project is driven by the question of political bias: under what conditions can LLMs be used democratically within forms of citizen participation that are them- selves still largely experimental? Addressing these socio-technical questions requires grounding in political theory and, more broadly, in philosophy-disciplines that provide the normative frameworks without which the democratic evaluation of AI systems cannot be mean- ingfully conducted.

Large language model (LLM) agents interact with external environments through tool invocation, but tool outputs can also expose them to indirect prompt injection (IPI) attacks. Existing defenses mainly rely on prompt hardening, content filtering, pre-generated plans, or permission constraints. These approaches often struggle with complex tasks or over-sanitize external content, making it difficult to balance security and utility. The key challenge is therefore to preserve execution flexibility while precisely identifying and removing the malicious content that actually induces unsafe actions. To address this challenge, we propose ActGuard, a pre-execution action auditing framework. Rather than judging whether external content is inherently suspicious, ActGuard assesses whether it causes the current action to deviate from a locally reasonable expectation. At each step, ActGuard predicts the tools likely to be used by the upcoming action and constructs a local tool prior without constraining the execution trajectory. Before execution, it compares the candidate action against this prior and performs tool-level contrastive analysis and parameter-level evidence localization to identify deviations in tool selection and action parameters. A verifier then examines the localized evidence, masks only spans confirmed as malicious, and regenerates the action from the sanitized context. This design preserves legitimate planning flexibility while minimizing information loss from indiscriminate filtering. We evaluate ActGuard on challenging benchmarks for tool-using agents. Results show that ActGuard reduces attack success rates to a level comparable to state-of-the-art defenses while maintaining task utility close to the no-attack setting, achieving a favorable security-utility trade-off. Our code is publicly available at: https://github.com/binzhwang/ActGuard.

Low-rank gradient projection reduces the optimizer-state memory cost of large language model (LLM) pretraining, but the steps and wall-clock time needed to reach a target quality remain a meaningful axis for improvement. We attribute this to two design choices in existing methods: the projection-rank budget is allocated uniformly across Transformer modules with heterogeneous projection sensitivity, and projecting a raw gradient attenuates its magnitude and direction jointly. We propose MoARa, which combines a static profiling-based module-aware projection-rank allocation with a block-wise magnitude-direction decomposition; the default block size is set in the neighborhood of the attention head dimension. Across five Transformer architectures spanning Llama, Qwen, and DeepSeek at 300M to 7B scales, GaLore with MoARa reaches standard GaLore's final perplexity in 37% fewer steps and 34% less wall-clock time on Llama 2 7B, with only 0.2% peak reserved memory overhead under standard graph compilation. Across the six low-rank pretraining methods we evaluate, module-aware rank allocation alone delivers directionally consistent step reductions on all six. On compatible hosts, the two-component design reaches up to 41.7% step reduction and 37.1% wall-clock reduction.

Large Reasoning Models (LRMs) pose a dual-surface safety challenge: both intermediate reasoning traces and final answers can contain harmful content. Existing alignment methods often operate at the whole-response level, allowing unsafe reasoning to be masked by a safe-looking final answer. We propose Segment-aware Listwise Target DPO (SaLT-DPO), which addresses this gap through three mechanisms: (1) segment-aware listwise alignment that decomposes responses into reasoning and answer segments, independently scores each segment's safety, and aligns length-normalized segment rewards with soft target distributions over multiple candidates; (2) joint safety coherence regularization that applies a weakest-link principle to promote safety consistency across both segments; and (3) utility anchoring on benign prompts to mitigate over-refusal and reasoning degradation. Experiments on three LRMs show that SaLT-DPO consistently reduces unsafe rates for both reasoning and answer segments while mitigating degradation in benign compliance and preserving general reasoning performance. Ablation studies demonstrate the complementary contributions of its components.

Language models can learn from experience, but raw solution trajectories are often too long and noisy to provide effective guidance. In this work, we propose Learning to Coach (L2C), a framework that trains a dedicated LLM-as-a-Coach to extract actionable experiential knowledge from an actor model's previous trajectory. The actor remains frozen, while the LLM-as-a-Coach is trained to maximize a reward given by the correctness of the actor's guided response. We study two such rewards: a same-instance reward, which improves subsequent responses on the original problem, and a cross-instance reward, which elicits knowledge that transfers to other instances. Across mathematical reasoning and interactive text-games, L2C consistently outperforms self-refinement and an untrained LLM-as-a-Coach. Running experiential learning for more iterations further improves accuracy and uses additional inference compute more effectively than enlarging the actor's decoding budget. The trained LLM-as-a-Coach also transfers to out-of-distribution tasks and adapts its guidance to the specific actor it coaches.

User prompts provided to large language models (LLMs) may contain sensitive or private information that can be misused by remotely deployed models, such as through inadvertent memorization during retraining. One way to protect user prompts is to execute the LLM inside a trusted execution environment (TEE), with the guarantee that the service provider has no access to computations performed within or information exchanged with the TEE. However, current TEEs are primarily CPU-based and significantly slower than GPUs optimized for LLM inference. To circumvent this, Tramer and Boneh (2019) proposed Slalom, which splits neural network inference between a TEE and an untrusted GPU and encrypts intermediate inputs sent to the GPU. We extend this split-inference architecture to LLM inference and instead protect intermediate inputs using differential privacy. We show that masking intermediate representations is necessary by showing that a prompt-reconstruction attack can recover prompts from these representations with nearly 80% accuracy. Our main contribution is a global sensitivity analysis of key LLM functions, which bounds the required scale of differentially private noise. Unlike encryption, differential privacy avoids quantization, allowing the LLM to remain in the floating-point domain. We also derive an upper bound on floating-point error from masking and noise cancellation in the TEE as a function of the privacy parameter epsilon. We implement our architecture using Intel TDX and evaluate it with two LLMs: Llama-3.2-3B and Qwen3-4B. Our split execution is nearly twice as fast as fully CPU-based inference inside TDX and 5-15 seconds faster than encryption-based Slalom while achieving higher accuracy. Finally, we demonstrate that prompt reconstruction, even with knowledge of the differential privacy mechanism, cannot recover more information than is contained in an unrelated prompt.

As large language models become capable translators of classical texts, a key challenge is deciding which outputs need expert review when no human reference exists. This study tests reference-free error triage through Pali-to-English translation. Three LLMs translated 15,493 passages. Five signals were compared: source novelty, source-candidate embedding distance, peer-translation disagreement, English-to-Pali backtranslation, and no-reference GEMBA scoring. Signals were calibrated on a 3,000-item reference-informed LLM-adjudicated sample and checked against a 500-item author-adjudicated anchor. Human references supported calibration and validation only; they were never used to compute the risk signals. Source novelty was a useful source-side risk prior but not a per-candidate error detector. Peer disagreement and backtranslation provided secondary signal. The strongest method was no-reference GEMBA scoring by a panel of models generally regarded as stronger than the translators: reviewing the top 10% by GEMBA risk captured 81.6% of panel-major errors in the calibration set. GEMBA also remained the best reference-free signal against the author anchor. A same-tier panel, with self-scoring excluded, remained useful but performed worse, indicating that evaluator strength matters beyond the prompt alone. A budgeted workflow is proposed, combining source novelty, peer disagreement, and stronger candidate-aware judging to allocate human review. Transfer to other classical languages, including Latin, Ancient Greek, and Sanskrit, remains to be tested.

Background. Large language models are increasingly used to help write biomedical text but may fabricate references to nonexistent work. How often large language models do so is not well characterized. Methods. We prompted 26 language models from eight developers (2023 to 2026) to supply a missing reference for each of 69 biomedical passages across ten domains. References were classified as verifiable (real paper with a resolving identifier), partial matches (real paper without a resolving identifier), fabricated (no matching indexed paper), or declined (the model refused to supply a reference). A reference was considered correct in every evaluated bibliographic field only when it was verifiable and its journal, year, and listed authors matched those of the cited paper. Results. Fabrication ranged from 10.2% (Claude Opus 4.8, which declined 52.1% of prompts) to 98.4% (Ministral 3B, which produced no verifiable reference). Claude Opus 4.6 and Claude Sonnet 4.5 produced similar proportions of verifiable references (77.6% and 76.6%) but named authors correctly in 78.7% and 28.7% of author-evaluable verifiable references, respectively, and were correct in every evaluated field in 54.6% and 19.9% of responses. GPT-5.5 was correct in every field in 48.1%. Across all models, 55.4% of responses were fabricated and 14.9% were correct in every field. Among the five tested models first released in 2026, the corresponding proportions were 35.3% and 31.8%, respectively. Conclusions. Fabrication remained common, and no model was correct in every evaluated bibliographic field in more than 54.6% of responses. Models that identify real papers may still misstate their metadata, so references produced with model assistance require verification before use.

Scaling large language models efficiently has motivated sparse capacity mechanisms such as Mixture-of-Experts and, more recently, conditional memory: token-indexed embedding tables that augment the backbone with cheap parametric lookups. Existing memory-embedding methods retrieve via a deterministic function of the surface form, which collapses different contextual senses of the same token (e.g., python the language vs. the animal) into a single fixed entry. We introduce Mixture of Memory Embeddings (MoME), a context-aware memory mechanism that replaces each token's single memory row with a mixture of M slots and uses a learned gate over the hidden state to choose which slots to read at each position. In controlled pretraining experiments across nanochat, Llama-3/MobileLLM, and Qwen3 backbones, MoME improves over Value Embedding, Bigram, and STEM baselines in iso-parameter and iso-training-FLOP settings, shows a more promising memory-size scaling trend at sub-billion scale, and remains efficient in training and inference. Qualitative routing analyses on polysemous tokens further suggest that the learned mixture exhibits a degree of semantic interpretability, dispatching the same surface token to distinct memory slots under different senses.

Large Audio-Language Models (LALMs) have substantially advanced general audio understanding, yet they remain limited in fine-grained temporal perception, particularly in precise event localization. Existing approaches primarily post-train LALMs to predict event boundaries as timestamp tokens. However, this generative formulation lacks explicit correspondence between the timestamp predictions and fine-grained acoustic evidence, limiting the precision and reliability of temporal localization. To address this issue, we augment the LALM with a dedicated frame-level grounding model while leveraging its semantic modeling capability to represent the event query. Specifically, the frozen LALM encodes the event query with audio as context, and the grounding model combines these query representations with fine-grained audio features to localize the target event at the frame level. Extensive experiments across diverse temporal grounding benchmarks demonstrate strong and consistent improvements over existing methods. Further evaluation shows that the grounding model can provide temporal evidence to support downstream reasoning.

Recently, the use of Large Language Models (LLMs) for different tasks in the Electronic Design Automation (EDA) life-cycle has been studied extensively, but an integrated view is lacking. Specifications are the foundation of this life-cycle, but they suffer from ambiguity when written in natural language, which especially affects the quality of LLM output. Formal specifications mitigate these ambiguities, but they come with their own challenges. On the other hand, Controlled Natural Language (CNL) specifications can serve as a middle-ground, reducing ambiguity while retaining interpretability. In this work, we propose an integrated view on the use of LLMs for EDA and establish an LLM-enabled behavior driven hardware development workflow. We introduce and define Formal Verification Gherkin Scenarios (FV Gherkin Scenarios), unlocking CNL specifications as the foundation for formally verified hardware designs via Formal Property Verification (FPV). Experimental evaluation shows that our workflow is able to outperform other established LLM-based methods by 2.48x in functional correctness of generated Register Transfer Level (RTL) designs and by 2.54x in formal coverage of generated assertions for FPV.

Discrete diffusion and flow models are a promising alternative to autoregressive language models, but compressing many-step sampling into fewer steps typically requires distilling a pretrained teacher model. This caps the student at the teacher's quality and requires a costly two-stage training pipeline. We introduce Discrete Beckmann Transport Models (DBTM), built on a time-independent flow whose autonomous transport map provably carries any point in the ambient space to a fixed point on the vertices of the simplex in a single step. We show that this fixed-point property is characterized by a conservation equation whose residual can be minimized directly from data, removing the requirement for a teacher flow and time conditioning. Under this construction, a partially trained map corresponds to the flow truncated at finite time, so generation reduces to iterating one map until it reaches a fixed point. We further extend the map to a partial-context interpolant where additional function evaluations act as refinement steps rather than ODE integration steps. On language modeling and reasoning tasks, DBTM enables one- and few-step generation that improves quality and accuracy over discrete diffusion and continuous flow baselines.

Through pre-training on extensive text and image datasets, current multi-modal large language models (MLLMs) achieve strong performance on general tasks. However, circuit schematics present a unique challenge for MLLMs due to their dense component layouts and distinct topological logic, demanding fine-grained structural parsing to extract the electrical semantics. To address this, we propose Circuit-MLLM, a multimodal reasoning framework that reformulates circuit topology analysis as a process of device localization, path tracing, and sequential reasoning within the latent space. We introduce a circuit knowledge mining mechanism that deeply aligns the model's latent representations with structurally rich features derived from multi-granularity circuit vision experts, enabling the model to effectively internalize topological semantics. Building upon these internalized semantics, we devise a topology-guided sequencing strategy that decouples reasoning from the rigid raster-scan order, enforcing stepwise inference along the circuit's topological logic in latent space. Across diverse circuit analysis tasks, Circuit-MLLM consistently outperforms strong baselines, notably achieving a 25% higher average score than GPT-5.1, which demonstrates the effectiveness of our framework in circuit schematic topology analysis. Code is publicly available at https://github.com/IC-Yuan/Circuit-MLLM.

Mechanistic interpretability has become the dominant lens for understanding frontier language models, as their inner workings are complex and inherently black boxes. To gain insights into these models, interpretable replacement networks (IRNs) are trained at all layers, exposing interpretable features through sparsely activated neurons. However, the faithfulness of an IRN is usually evaluated only empirically on clean data, and we show that even semantically minor input perturbations flip the dominant IRN features-and thus the human-understandable interpretation-across five open-weight model families (GPT-2 small, Gemma 2 2B, Gemma 3 1B, Llama 3.2 1B, R1-Distill-Qwen 1.5B). We propose the first formal verification framework for the faithfulness of an IRN, where reachability analysis certifies a sound upper bound of the faithfulness gap in adversarial scenarios. Moreover, we show that verification-aware training of IRNs substantially tightens this certified bound, restoring a feature-level interpretation that safety auditors can act on. Together, these results give, to the best of our knowledge, the first formal guarantees for mechanistic interpretability of large language models.

When a direct proof of a statement $S$ seems hard or even impossible to obtain, there may exist another statement (or set of statements) $S^{*}$, somehow related to $S$, on the basis of which $S$ can be proved. In order to investigate what options can be used to move from $S$ to $S^{*}$, four kinds of semiotic relations inspired by the four master tropes of semiotic research are briefly reviewed. Specifically, our syntagmatic, paradigmatic, antithetic and meronymic relations correspond, respectively, to metonymy, metaphor, irony and synecdoche. It is suggested that these four semiotic relations determine the options to move from $S$ to $S^{*}$, leading to proof by inference, proof by analogy, proof by contradiction, and proof by case analysis. To examine how the four relations are actually used across different kinds of argument, we complement the framework with an empirical study. We turn the four relations into explicit operational definitions and apply them to a cross-genre corpus of mathematical, legal, and everyday argument using a panel of large language models. We find that the relations are used very unevenly across genres: mathematical proofs draw on all four, whereas legal and everyday reasoning rely almost entirely on inference.

The reasoning ability of large language models (LLMs) is a critical factor for practical LLM-based applications. To investigate the current reasoning capability of LLMs, we clarify the types of errors that arise in LLMs' reasoning processes on mathematical datasets. We focus on problems where LLMs produce an incorrect answer. We define errors in the reasoning process as reasoning errors and manually analyze the features of reasoning errors. We defined and classified 21 error classes and identified the frequently occurring classes among them. Beyond qualitative evaluation, we leverage the evaluation results to improve the reasoning capability. We designed a prompt that explicitly focuses on eight error classes. The experiments demonstrate that this prompt effectively improves reasoning performance. Furthermore, the results suggest that the frequent reasoning errors identified in this paper are common across LLMs of comparable scale.

Entrepreneurial cognition is a foundation of entrepreneurship research. Yet the growing involvement of large language models (LLMs) in entrepreneurial work extends the cognition question beyond human actors to systems whose internal representations remain largely unexplored. We introduce artificial entrepreneurial cognition, the functional organisation of entrepreneurship-relevant representations and computations inside artificial intelligence (AI) systems. We bring mechanistic interpretability into entrepreneurship research through representation engineering. Focusing on opportunity recognition (OR), we construct 636 matched OR-present and OR-absent scenario pairs and recover an OR direction in Llama 3.1 8B-Instruct. Rather than infer the construct from outputs, we intervene directly on this direction, steering the model up and down along what we call the opportunity recognition dial, and its opportunity judgments shift with it. To our knowledge, this is the first causal intervention on an internal representation of an entrepreneurship construct inside an LLM. Held-out tests, lexical and topical controls, behavioural ablation, and geometric comparisons show that the direction is recoverable, consequential, and distinct from the opportunity evaluation and exploitation directions, although steering it also shifts judgments about these neighbouring stages. Recovery, signed steering, and geometric separation hold across four additional LLMs spanning different scales and families. These results give the contested distinction between opportunity recognition and evaluation a concrete representational form inside AI systems. More broadly, they establish internal representations as a new object of entrepreneurship inquiry and show how entrepreneurship theory can guide their identification, causal manipulation, and interpretation.

Span-level masked language modeling (MLM) has shown to be advantageous to pre-trained language models over the original single-token MLM, as entities/phrases and their dependencies are critical to language understanding. Previous works only consider span length with some discrete distributions, while the dependencies among spans are ignored, i.e., assuming that the positions of masked spans are uniformly distributed. In this paper, we present POSPAN, a general framework to allow diverse position-constrained span masking strategies via the combination of span length distribution and position constraint distribution, which unifies all existing span-level masking methods. To verify the effectiveness of POSPAN in pre-training, we evaluate it on the datasets from several NLU benchmarks. Experimental results indicate that the position constraint is capable of enhancing span-level masking broadly, and our best POSPAN setting consistently outperforms its span-length-only counterparts and vanilla MLM. We also conduct theoretical analysis for the position constraint in masked language models to shed light on the reason why POSPAN works well, demonstrating the rationality and necessity of POSPAN.

Test-time compute has emerged as a major approach to improving the capabilities of Large Language Models (LLMs). However, existing test-time reasoning paradigms rely heavily on externally imposed control, either through fixed reasoning programs or through costly expansion in constrained search spaces, limiting both generalization and efficiency. We propose State of Thought (SoT), a new reasoning paradigm that enables endogenous reasoning in LLMs, with the model's internal reasoning state governing how reasoning unfolds. Concretely, SoT extracts a compact dynamics-geometric state from the model's internal information transfer and uses a 582-parameter controller on frozen backbones to selectively activate historical reasoning support useful under the current reasoning state, framing reasoning as a state-conditioned process over evidence rather than an externally prescribed token chain. Across quantitative (1.34x), general (1.62x), symbolic-and-code (1.76x), and long-context (2.51x) reasoning on 3 LLMs and 16 datasets, SoT consistently improves mean-baseline accuracy while reducing generated tokens by 62.6% and end-to-end latency by 44.6%. Across 2 VLM scales and 3 reasoning tasks, it improves mean accuracy by 3.8 points over reasoning baselines, with 74.9% fewer completion tokens and 73.5% lower latency than search-based methods. Under constrained access, SoT retains 38.2%/36.5% mean accuracy gains in training-free/embedding-only settings, while trajectory-only judging reaches 84.1% agreement across 3 API models. Together, endogenous state-driven reasoning provides a generalizable and efficient alternative.

Large language models increasingly analyze computational results and draft manuscripts, making reliable communication as important as correct analysis. Using fixed computational evidence, we tested whether assigning comparisons across modeling choices elsewhere in a research workflow changes manuscript reporting. In constrained sentence-writing tasks, Anthropic's Claude Sonnet 5 often omitted numerical qualifications when detailed comparisons were assigned to a group repository, but retained them more often when the same comparison was assigned to Supporting Information or its own working notes; Claude Opus 5 was less sensitive. These effects did not follow a simple accessibility ordering. A targeted placement rule largely restored sentence-level qualification, whereas a generic accuracy reminder did not. Longer contributions retained numerical qualifications, although some summaries across computational settings were still redirected to the repository. Thus, documenting context within an AI workflow does not ensure its communication where readers encounter the result.

Real-time pedestrian detection in driving scenes is constrained by three coupled failure modes: tiny targets lose discriminative evidence, occlusion weakens geometric support, and weather or illumination changes distort appearance statistics. We formulate the detector through a unified contour-guided spectral routing view rather than treating frequency processing, attention, and boundary reasoning as independent add-ons. The detector routes information in a prescribed order: spatial evidence is first augmented with global spectral context, deep representations then exchange spatial and spectral cues, and cross-scale fusion is finally conditioned on boundary--semantic disagreement. This ordering yields a compact representation pipeline in which low-frequency context stabilizes global structure while high-frequency evidence protects small-object contours. We further retain a wavelet-subband training transformation that perturbs low- and high-frequency coefficients independently, targeting appearance shifts caused by fog, rain, snow, and low illumination. The formulation exposes a single routing variable at each stage and distinguishes reusable signal transforms from the task-specific policy that decides where each signal is injected. On CityPersons, the proposed detector obtains 70.4 AP$_{50}$ and 44.2 AP$_{50:95}$, compared with 68.1 and 42.2 for RT-DETR, while the full wavelet-augmented configuration reaches 71.1 and 44.6.

Multi-turn interactions with LLMs are becoming increasingly common in information-seeking scenarios. However, user queries are often ambiguous and context-dependent, making them ill-suited for direct use as retriever queries. Conversational query reformulation (CQR) addresses this issue by rewriting the current utterance into a stand-alone query grounded in the dialogue history. Recent LLM-based CQR approaches achieve strong performance; however, their repeated LLM invocations and misalignment with downstream retrievers remain challenges. In this work, we begin from the observation that retrievers are highly sensitive to content ordering: simply reordering the same content can lead to changes in retrieval coverage and performance. Based on this, we propose a novel training-free method that generates a very large number of queries with minimal LLM usage by compositionally combining a small set of atomic components. We further apply LLM reasoning to construct a high-quality document set that balances precision and recall while capturing the user's core intent. Our framework generalizes across both open- and closed-source LLMs as well as dense and sparse retrievers. It achieves strong performance on four widely used conversational benchmarks, with up to 22.5% relative MRR improvement over the previous state-of-the-art baseline with far fewer LLM calls.

Causal claims about large language model (LLM) internals rest on measurements. Those might include a projection, a cosine, an ablation delta, or an interchange patch among others. These measurements fail in specific, diagnosable ways that return a plausible number instead of an error, so a broken instrument can easily read as a finding. A covariance-matched null can saturate until every direction looks typical, a per-head attribution can overshoot the true residual write threefold on reordered-normalization architectures, an interchange patch can go sign-chaotic because its outcome is pinned at a ceiling, or a read-from verdict can be an artifact of measuring past the layer where the model already decided. This note documents six such failures from a causal interpretability program on refusal and moral representation, spanning several papers and a four-model open-weight panel; each mode is established on one or two of the four. For each we give the tell that catches it and a protocol keyed to a detectable trigger (reordered normalization, massive activations, a low-dimensional decision channel), so we and readers can check whether a given setup is exposed. The discipline reduces to four moves: calibrate against a positive-control ladder, certify with an orthogonal cell, compute power before spending compute, and state every read-from verdict at a depth referenced to the model's commitment. The evidence is four architectures across three families within a single program; external replication across programs is future work.

We introduce Enemray, a Hassaniya-centric language model that enables general-purpose interaction in Hassaniya. Enemray is trained around a stability--plasticity objective: acquire strong Hassaniya linguistic and cultural competence while preserving the general reasoning, multilingual, instruction-following, and safety behaviors of a capable instruction-tuned model. The development pipeline separates language acquisition from behavioral specialization. A separately assembled continual-pretraining corpus provides broad exposure to natural Hassaniya and Mauritanian text; layer-selective continual pretraining learns a compact language-specific parameter update; that update is transferred into the instruction-tuned parameter space; and supervised post-training develops conversational, cultural, literary, task-oriented, and cross-lingual behavior. The supervised corpus integrates selected public Hassaniya and Mauritanian resources with a substantially larger body of newly collected, reconstructed, curated, and constructed instruction data, while policy-generated replay provides a retention signal from the reference model's own behavior distribution. The resulting collection is substantially larger and broader in purpose than existing Hassaniya text resources. In evaluation, Enemray achieves the strongest English to Hassaniya translation among the compared open and proprietary models and the highest overall score on Mauritanian translation error detection, while retaining most of the general capabilities of its instruction-tuned base model on mathematical reasoning, knowledge, code generation, and function calling. This report describes the motivation, data construction, model design, training methodology, and evaluation of Enemray.

Performing deliberate mathematical reasoning in visual contexts is a hallmark of advanced Multimodal Large Language Models (MLLMs) and requires a sophisticated synthesis of perceptual grounding and symbolic logic. However, in the realm of mathematical functions, our investigation reveals a critical modality interference phenomenon: even advanced models, while performing textual computational reasoning, tend to disregard or misinterpret essential visual cues. To address this challenge, we propose Func-R1, which synergistically harmonizes precise visual perception and rigorous logical reasoning. Concretely, built upon an explicitly decoupled architecture, we employ a hierarchical post-training framework to progressively identify critical visual evidence and conduct in-depth theoretical reasoning. Furthermore, the Perception-Aligned Theoretic Optimization (PATO) strategy is proposed to steer policy updating towards internalizing fundamental theoretical properties while dynamically rectifying heterogeneous visual information throughout the reasoning process. Extensive experiments across diverse benchmarks demonstrate that Func-R1 delivers the optimal performance among open-source MLLMs, even surpassing GPT-5 with an 8.4% improvement on MathVerse's function-oriented tasks.

Recent advances in speech synthesis and voice conversion have made deepfake speech increasingly realistic, making generalization to unseen spoofing attacks a critical challenge. Pretrained speech and audio models offer a promising direction for improving robustness to such unseen attacks. Self-supervised learning (SSL) models capture fine-grained, low-level acoustic characteristics, whereas Auditory Large Language Models (ALLMs) provide higher-level contextual representations. These complementary views can provide useful cues for improving generalization to unseen attacks. However, direct fusion does not explicitly disentangle information shared across the two views from view-specific complementary information, limiting effective cross-view integration. To address this, we propose CRAF, a cross-view residual-aware fusion framework that uses ALLM-guided cross-view attention to enrich SSL representations and adopts ALLM as a high-level reference to separate ALLM-explainable information from complementary SSL residual information. The residual is selectively refined through adaptive gating and integrated through SSL-primary fusion. Experiments on ASVspoof 5 show that CRAF with Kimi-Audio achieves an EER of 5.96% and a minDCF of 0.1192, demonstrating robustness to unseen spoofing attacks.

The CRITICS project addresses science accessibility and literacy by converging advanced Machine Translation (MT) based on Large Language Models (LLMs) with educational technology. By leveraging MT systems specifically optimized for scientific content, educational institutions can provide accurate, culturally relevant translations of scientific materials in students' native languages, ensuring that complex scientific concepts are comprehensible while maintaining technical accuracy. Building on these translations, the project explores the design and evaluation of innovative science teaching-learning proposals grounded in curriculum-aligned teaching-learning. Thus, CRITICS will investigate key components of scientific argumentation and critical thinking practices together with textual feedback aligned with learning objectives and assessment criteria inspired by competence-based evaluation frameworks. CRITICS aims to break down language barriers to accessing cutting-edge research and educational materials currently available only in high-resourced languages, thereby facilitating the democratization of scientific knowledge and fostering critical thinking in science education.

Detecting harmful memes is critical for maintaining safe online communities. However, harmful intent is often implicit, arising from visual-textual incongruity and cultural stereotypes, which challenges existing multimodal detectors. We propose SyRHM, a framework that decomposes harmful meme detection into meaning-grounded retrieval and symbolic-language-enhanced multi-stage reasoning. SyRHM retrieves semantically related memes by parsing multimodal content into textual elements and descriptions, providing grounded context beyond surface-level similarity. Building on the retrieved context, SyRHM uses a translator stage to convert multimodal inputs into symbolic intermediate representations, and then performs multi-stage reasoning via planner and solver stages, enabling expressive and interpretable analysis of harmful intent. Experiments on FHM, HarM, and MultiOff demonstrate the effectiveness of SyRHM, achieving superior performance on most evaluation settings against multimodal and reasoning-based baselines, while providing reasoning traces for harmful content. The code is available at: https://github.com/Scabbards1500/SyRHM

Large language models (LLMs) are increasingly deployed with external tools that extend what they can do beyond their own knowledge. Tools help on tasks that need external information, but their availability may also change how a model handles questions that do not need them. Prior work has mostly asked whether models select and use tools appropriately; whether an unnecessary tool changes the correctness of answers has received less attention. We ask whether making a related but unnecessary tool available affects a model's ability to answer from its own knowledge, and whether a preceding tool interaction changes this behaviour. We construct 500 query pairs across 10 knowledge domains. Each pair consists of a tool query, which needs the domain's tool, and a closed-domain query, which does not. Six LLMs are evaluated with the tool unavailable, available, and available after a prior tool call. Across 3,000 baseline trials the pooled answer rate is 98.2%. When an unnecessary tool is available it falls to 63.5%, with large differences between models. The decrease occurs even when the tool is rarely called, so it cannot be explained by unnecessary tool invocation alone. A one-sentence scope-aware system instruction recovers most of the lost answers.