Search research

3D visual grounding aims to localize the target object in a 3D scene from a natural language query, requiring both fine-grained semantic understanding and viewpoint-dependent spatial reasoning. Existing methods typically formulate semantic understanding as an auxiliary closed-set object classification task and rely on multi-view feature aggregation for viewpoint reasoning, limiting semantic generalization and weakening viewpoint-specific evidence. We observe that vision-language models naturally provide complementary capabilities through open-vocabulary semantic understanding and global scene perception. Based on this insight, we propose GuideGround, a VLM-guided framework that complements rather than replaces task-specific grounding models by leveraging VLMs for semantic enhancement and viewpoint-specific hypothesis verification. Specifically, we replace auxiliary closed-set object classification with VLM-generated object semantic descriptions to enhance semantic understanding. Meanwhile, instead of directly aggregating multi-view representations, we preserve viewpoint-specific grounding hypotheses through per-view grounding and explicitly verify them using VLMs across candidate viewpoints. Extensive experiments on the ReferIt3D benchmark demonstrate that GuideGround consistently outperforms previous state-of-the-art methods. Comprehensive ablation studies further confirm the effectiveness of both the proposed semantic understanding and viewpoint reasoning strategies.

Information locality, the tendency for syntactically related words to appear close together, shapes both human language processing and language model learning. While prior work has examined whether language models can acquire impossible languages, it remains unclear whether they can recover natural language from such input and what this reveals about their inductive biases. We address this by complementing learnability-based approaches with a reconstruction framework: fine-tuning GPT-2 models pre-trained on impossible languages to reconstruct natural English from three perturbation types. Our findings show that the recovered structures exhibit shorter dependency lengths than the original text, mirroring the locality preference observed in unconstrained language model generation and providing a quantitative signature of an architectural bias that learnability experiments alone do not reveal. Recovery difficulty increases with the degree of locality disruption. Structural recovery (dependency Triple F1) dissociates from surface recovery (Exact Match), while fluency dissociates from faithful reconstruction under global shuffling. Sentence length further modulates performance: longer sentences facilitate recovery when local structure is preserved but lead to complete collapse under global shuffling. Finally, recovery difficulty tracks learnability difficulty across perturbation types, suggesting that information locality is the shared constraint governing both.

Language Identification (LID) is an important component of many multilingual natural language processing pipelines, where it facilitates corpus curation, training data analysis, and cross-lingual evaluation of large language models. Despite near-perfect performance on high-resource languages, existing systems remain brittle in low-resource and closely related language settings. We introduce UniLID, a simple and efficient LID method based on the UnigramLM tokenization algorithm. In short, to predict a string's language label, we simply ask: under which language's unigram distribution is this string most likely? Our formulation is data- and compute-efficient, supports incremental addition of new languages without retraining existing models, and can naturally be integrated into existing language model tokenization pipelines. Empirical evaluations against widely used baselines, including fasttext, GlotLID-M, and CLD3, show that UniLID achieves competitive performance on standard benchmarks, reaches 69% accuracy with five labeled samples per language and 89% with 25, and delivers large gains on fine-grained dialect identification.

Computing legal indicators from normative texts is a key task in legal monitoring and policy evaluation, but presents significant challenges due to the complexity, scale, and interpretive nature of legal language, as well as the variability in available document quality. Existing natural language processing techniques and generative models can assist in legal analysis, but often suffer from high risk of hallucinations and lack the interpretability and evidence grounding required for reliable indicator computation. This paper presents N2I-RAG (From Norms to Indicators), an agentic retrieval-augmented generation framework designed to automate the computation of legal indicators in a transparent and traceable way. We integrate adaptive retrieval, llm-based agents, and validation mechanisms in a modular pipeline, where each component performs a defined role in filtering, retrieving, and assessing evidence, and in producing binary legal outcomes linked to identifiable legal provisions. The framework emphasizes traceability by requiring explicit explanations of intermediate decisions and final indicator assignments. We evaluate N2I-RAG using an in-house constructed French marine environmental law corpus that includes both scanned and digital sources. Comparative experiments with multiple language model families demonstrate that the proposed approach consistently outperforms baseline systems, and generalizes well when tested on 2 different bans. The results indicate that agentic retrieval-augmented generation can bridge open-text legal language and standardized indicator computation, offering a foundation for transparent and scalable legal observatories.

Single-cell representation learning (SCRL) from gene expression data offers a way to uncover the complex regulatory logic underlying cellular function. Inspired by large language models in natural language modeling, several single-cell pretrained models have recently been proposed that treat genes as tokens and cells as sentences. However, these models are fundamentally limited by the long-tailed nature of cell-type distributions and struggle to generalize under covariate shifts in gene expression data. While fine-tuning is often used to mitigate these issues, we observe that performance remains bounded. To address this challenge, we introduce CellRefine, a post-pretraining method that operates between the pretraining and fine-tuning stages of a single-cell foundation model. CellRefine uses a multi-faceted objective that incorporates marker-gene sets as structural priors to guide post-pretraining and refine the latent embedding manifold of cells. Across multiple computational biology tasks, empirical results show that CellRefine consistently improves downstream performance, yielding gains up to 15%.

Natural language is an intuitive interface for humanoid robots, yet streaming whole-body control requires control representations that are executable now and anticipatory of future physical transitions. Existing language-conditioned humanoid systems typically generate kinematic references that a low-level tracker must repair reactively, or use latent/action policies whose outputs do not explicitly encode upcoming contact changes, support transfers, and balance preparation. We propose \textbf{DAJI} (\emph{Dynamics-Aligned Joint Intent}), a hierarchical framework that learns an anticipatory joint-intent interface between language generation and closed-loop control. DAJI-Act distills a future-aware teacher into a deployable diffusion action policy through student-driven rollouts, while DAJI-Flow autoregressively generates future intent chunks from language and intent history. Experiments show that DAJI achieves strong results in anticipatory latent learning, single-instruction generation, and streaming instruction following, reaching 94.42% rollout success on HumanML3D-style generation and 0.152 subsequence FID on BABEL.

Assistive autonomous robots operating in multi-agent environments require efficient strategies to locate specific individuals among multiple residents. Current social navigation methods focus on reactive collision avoidance and trajectory adaptation, but lack mechanisms to proactively gather information through human-robot communication. We introduce Communication-enabled Social Navigation (CommNav). In this novel task, robotic agents actively seek assistance from residents to locate target individuals by requesting information about recent sightings, locations, and movements. To evaluate CommNav, we extend Habitat 3.0 to create Habitat 3.0c, a communication-enabled variant supporting multi-human environments with information exchange protocols. Adding our communication module (COMM) to a state-of-the-art social navigation model yields a 10 percentage-point improvement in Episode Success. We further investigate the transition from structured data to natural language by evaluating models trained on LLM-generated instructions and on colloquial instructions collected from a human study. Our experiments reveal that: (i) explicit human-robot communication substantially enhances multi-person navigation performance; (ii) pre-training COMM on a communication pretext task effectively addresses the challenge of occasional interaction signals; and (iii) the navigation policy is highly robust to natural, colloquial human language, achieving an episode success statistically similar to the model using perfect structured data.

Inspired by recent breakthroughs in large language models for natural language processing, foundation models have emerged as a promising paradigm for zero-shot time series forecasting, enabling accurate predictions on datasets never seen during pre-training. Ranging from tens to hundreds of millions of parameters, these models are pre-trained on vast and diverse collections of time series, learning generalizable representations that support both point and probabilistic forecasting. This approach alleviates the need for dataset-specific model design and manual tuning, offering a unified solution across forecasting problems. In this work, we review the main architectures, pre-training strategies, and optimization methods underpinning these models. We further investigate post-pre-training fine-tuning of selected foundation models to enhance their performance on specific datasets. Our empirical results demonstrate that this step consistently improves forecasting accuracy over the zero-shot baseline.

Tool and agent routing from natural-language prompts is naturally a set-valued prediction problem: a single query may require multiple agents, while over-selection increases execution cost. The benchmark introduced here is derived from WildChat and contains 3,000 prompts over a fixed 12-agent catalog, with AI-assisted heuristic labels under a fixed schema and controlled rebalancing for multi-label evaluation. The evaluation protocol combines set-level metrics (Precision, Recall, F1, Jaccard, and Exact Match), latency, an execution-oriented capability-coverage simulation, and a constrained weighted-routing setting based on ordinal agent-cost tiers. Compared methods include nearest-neighbor matching, linear multilabel classification, dependency-aware baselines, a fine-tuned encoder, deterministic weighted post-scoring via Weighted Agent Routing (WAR), and a zero-shot LLM baseline. Results show that supervised routers substantially outperform nearest-neighbor and zero-shot LLM routing. The fine-tuned encoder achieves the strongest unconstrained set accuracy, while the linear multilabel model provides the strongest practical baseline. In the constrained setting, the weighted routing layer improves utility when applied on top of strong supervised scorers, with the largest gain observed for Encoder+WAR. Overall, the benchmark and evaluation protocol support reproducible study of accuracy-cost trade-offs in fixed-catalog multi-agent routing.

More than a billion people communicate in Indic languages, yet the natural language processing infrastructure serving them remains fragmented and underdeveloped. The cause is structural: the field organizes its tools and benchmarks around individual languages or small subsets of genealogical language families, building separate analyzers, parsers, and datasets for each language and starting over for the next. This overlooks a deep regularity. Through more than two millennia of convergence around Sanskrit, Indic languages came to share a morphosyntactic architecture formalized in Pānini's grammar, the Astādhyāyī. This cuts across genealogical lines, uniting languages through a common framework. We argue that this Pāninian framework supplies a unifying computational architecture the field has lacked, and that benchmarks grounded explicitly in it would make Indic language systems more accurate, more data-efficient, and more transferable, effectively merging many apparently disparate and sparse Indic language resources into a single high-resource metalanguage bedrock. We propose a four-part benchmark suite to render this shared architecture explicit, measurable, and ready to be leveraged for practical applications. Moreover, we underscore the question it raises for interpretability research: whether neural models trained on these languages come to represent Pānini's categories on their own.

With the advent of Large Language Models, single-task and token-based multi-task models have evolved into instruction-based systems that infer task and target language implicitly from natural language prompts. This trend is reflected in IWSLT's Instruction Following Track, which this year introduced new tasks including an unknown surprise task, posing a genuine challenge against overfitting to known tasks. We present KIT's submission to the Long and Short Instruction Following tracks in the unconstrained setting. Our approach combines a general data augmentation pipeline that converts short-form corpora into long-form training data through segment concatenation, LLM-based label generation, and cross-lingual translation, yielding over 1M instances across six tasks and four languages. We further show that likelihood-based re-ranking, while highly effective for ASR, systematically degrades semantic tasks by spuriously selecting candidates generated from segmented audio processing rather than holistic long-form inference, a failure mode resolved by combining likelihood with Minimum Bayes Risk decoding.

Sign Language Production (SLP) is the task of generating avatar sign language motion from natural language text. The quality of the generated motion is typically evaluated by a motion-space Fréchet distance (FID) and back-translation (BT) BLEU score on benchmarks such as How2Sign. Both metrics can improve substantially while the underlying generator fails to faithfully represent the sign language gestures. In this work we propose to evaluate the generated motion at three independent levels: ($\tau1$) initial-pose conditioning, ($\tau2$) output diversity, and ($\tau3$) target faithfulness. We compute these as pairwise-distance ratios using latent representations of a frozen motion autoencoder (MoAE). We evaluate 14 SLP model checkpoints on the How2Sign dataset, including a re-implemented Neural Sign Actors (NSA), and show that $\tau3$ faithfulness is never attained, while FID varies by nearly two orders of magnitude and is uncorrelated with faithfulness. We show that on the isolated gloss dataset ASL3DWord favorable $\tau3$ can be attained, hence isolating the size of the sentence-level paired-dataset as the bottleneck.

The scaling exponent $α$ in neural scaling laws $L(N) \propto N^{-α}$ is commonly treated as a fixed constant set by architecture and data. We present evidence that $α$ depends systematically on the optimizer. In controlled random-feature regression experiments -- the canonical theoretical framework for neural scaling -- we measure $α$ across five optimizer variants and six spectral conditions. Preconditioned optimizers consistently yield steeper scaling (larger $α$), with the $α$-shift increasing across most of the tested spectral range, peaking near $s = 1.5$, and remaining large at $s = 2.0$. At $s \approx 1.0$ (characteristic of natural language), the full natural gradient achieves $α\approx 0.31$ versus $α\approx 0.12$ for gradient descent -- a $2.6\times$ larger fitted exponent that, within the random-feature model, compounds with each model-size doubling. Whether and how this exponent shift transfers to large-scale LLM training -- where recent evidence suggests the advantage may attenuate with scale -- remains an important open question. Our results imply that scaling-law forecasts should account for optimizer choice, and we provide a spectral diagnostic predicting when advanced optimizers will pay off.

Tool-integrated reasoning (TIR) has emerged as a dominant paradigm for mathematical problem solving in language models, combining natural language (NL) reasoning with code execution. However, this interleaved setup has three key limitations: code often acts as a post-hoc verifier, intermediate NL computations are error-prone, and NL and code play overlapping rather than clearly distinct roles. We propose ThinC (Thinking in Code), a framework in which code itself serves as the reasoner rather than as a tool invoked by NL. A ThinC trajectory begins with a brief NL planning step, after which all reasoning unfolds through code blocks connected only by their execution outputs. We distill 12.2k code-centric trajectories from a teacher model and train ThinC-1.7B and ThinC-4B with supervised fine-tuning followed by reinforcement learning. ThinC-4B consistently outperforms every TIR baseline on five competition-level math benchmarks and even surpasses the much larger Qwen3-235B-A22B-Thinking. Further analysis shows that ThinC reasons through code: 99.2% of its final answers are grounded in interpreter output, and the model recovers reliably from code execution failures without intermediate NL reasoning. Our code and models will be released soon.

The transformer is the most popular neural architecture for language modeling. The cornerstone of the transformer is its global attention mechanism, which lets the model aggregate information from all preceding tokens before generating the next token. One common variant of attention is called local attention, which restricts each token to aggregating information from a bounded window of predecessors, reducing the quadratic cost of global attention to linear. Although this restriction is usually motivated by efficiency, it has also been found to improve model quality, a phenomenon that has so far lacked a satisfactory explanation. We provide a formal account of this phenomenon in terms of recognizer expressivity. It has been shown that fixed-precision transformers with global attention correspond to a fragment of linear temporal logic containing a single past operator. We additionally prove that adding local attention introduces a second temporal operator, strictly enlarging the class of recognizable regular languages. Moreover, global and local attention are expressively complementary: neither subsumes the other, and combining them yields the richest fragment. Experiments on formal language recognition and natural language modeling corroborate the theory, showing that hybrid global--local transformers outperform their global-only counterparts.

Three generations of software have transformed the role of artificial intelligence in society. In the first, programmers wrote explicit logic. In the second, neural networks learned programs from data. In the third, large language models turn natural language itself into a programming interface. These shifts reach far beyond computer science, reshaping how societies generate knowledge, make decisions, and govern themselves. While generative adversarial networks introduced the era of deepfakes and synthetic media, large language models have added a new class of systemic risks. This report performs "mindhunting" for LLMs by applying a forensic-psychology profiling methodology to characterize AI based on documented features, e.g., hallucinations, bias and toxicity, sycophancy, fabrication and confabulation, knowledge without understanding, discontinuity and the inability to learn from experience, jagged intelligence, shortcuts and fractured representations. The resulting profile reveals an "entity" that confabulates fluently, amplifies its users' biases, possesses encyclopedic recall without causal understanding, and erodes the competence of those who depend on it. The implications extend to institutional erosion across law, academia, journalism, and democratic governance. To address these challenges, this report proposes a four-pillar framework for AI resilience: (i) cognitive sovereignty, which preserves the capacity for independent judgment, (ii) measurable control, which translates ethical commitments into enforceable standards and red lines, (iii) partial autonomy, which maintains human agency at critical decision points, and (iv) openness to guarantee transparency and accessibility (open-source, open-access, and open-data). This report is an updated and extended version of arXiv:1912.08786v1.

Precise object placement remains underexplored in aerial manipulation, where most systems rely on predefined target coordinates and focus primarily on grasping and control. Specifying exact placement poses, however, is cumbersome in real-world settings, where users naturally communicate goals through language. In this work, we present AeroPlace-Flow, a training-free framework for language-grounded aerial object placement that unifies visual foresight with explicit 3D geometric reasoning and object flow. Given RGB-D observations of the object and the placement scene, along with a natural language instruction, AeroPlace-Flow first synthesizes a task-complete goal image using image editing models. The imagined configuration is then grounded into metric 3D space through depth alignment and object-centric reasoning, enabling the inference of a collision-aware object flow that transports the grasped object to a language and contact-consistent placement configuration. The resulting motion is executed via standard trajectory tracking for an aerial manipulator. AeroPlace-Flow produces executable placement targets without requiring predefined poses or task-specific training. We validate our approach through extensive simulation and real-world experiments, demonstrating reliable language-conditioned placement across diverse aerial scenarios with an average success rate of 75% on hardware.

Teaching machines to emulate natural handwriting styles remains an open challenge, as it requires synthesizing stroke sequences that dynamically vary in shape, texture, pressure and script - not only across individuals, but also within a single person's handwriting. Attempts at this challenge have largely explored deep learning methods in both online and offline settings. However, these approaches are often constrained by style-specific architectural choices, heavy reliance on large datasets, high compute costs, and a lack of flexible control over writing styles through natural language. To this end, we introduce HandwritingAgent, a language-driven agent that can synthesize natural handwriting sequences directly in Scalable Vector Graphics (SVG) format with no need for style-specific training. The agent leverages a large reasoning model to geometrically analyse and autoregressively generate target handwritten glyphs as stroke sequences in a discrete grid canvas environment. Generation is conditioned on texts provided in either conversational or non-conversational mode, along with a reference handwriting-style image. Experiments on diverse handwriting tasks spanning imitation, recognition, multi-lingual handwriting synthesis, and generation of complex handwritten maths and science expressions indicate substantial improvement in performance, with HandwritingAgent matching or surpassing state-of-the-art generative handwriting models, while providing a more efficient, controllable, and generalizable synthesis method.

Natural language interfaces to structured databases are becoming increasingly common, largely due to advances in large language models (LLMs) that enable users to query data using conversational input rather than formal query languages such as SQL. While this paradigm significantly improves usability and accessibility, it introduces new security risks, particularly the amplification of SQL injection vulnerabilities through the prompt-to-SQL translation process. Malicious users can exploit these mechanisms by crafting adversarial prompts that manipulate model behavior and generate unsafe queries. In this work, we propose a multi-layered security framework designed to detect and mitigate LLM-mediated SQL injection attacks. The framework integrates a front-end security shield for prompt sanitization, an advanced threat detection model for behavioral and semantic anomaly identification, and a signature-based control layer for known attack patterns. We evaluate the proposed framework under diverse and realistic attack scenarios, including prompt injection, obfuscated SQL payloads, and context-manipulation attacks. To ensure robustness, we generate and curate a comprehensive benchmark dataset of adversarial prompts and assess performance across a fine-tuned LLM configuration. Experimental results demonstrate that the proposed approach achieves high detection accuracy while maintaining low false-positive rates, significantly improving the secure deployment of LLM-powered database applications.

Diabetic Retinopathy (DR) is an aggressive retinal disease and a leading cause of global blindness, yet its clinical management is currently hindered by the black-box nature of diagnostic AI. While deep learning models achieve high classification accuracy, there is a critical lack of explainability methods capable of detailing the exact anatomical landmarks and lesion distributions that lead to a clinical decision for DR. Therefore, we propose HSQ-VLM, a novel quadrant segmentation pipeline on fundus images that utilizes a Landmark-Anchored Cartesian Cross-Attention mechanism to unify visual feature extraction with structured clinical reasoning. Unlike traditional methods that rely on arbitrary image partitioning, our pipeline implements 4-quadrant Topological Latent Partitioning (TLP) to dynamically align retinal features with a fovea-centered coordinate system. This allows the Vision-Language Model to generate natural language reports that quantify pathology with anatomical precision. On a dataset of 3,500 high-resolution fundus images, this innovative methodology achieved a lesion detection sensitivity of 99.6% for hemorrhages and 96.4% for microaneurysms, while demonstrating a significant reduction in boundary-ambiguity errors compared to standard segmentation baselines.

Autonomous driving has progressed from modular pipelines toward end-to-end unification, and Vision-Language-Action (VLA) models are a natural extension of this journey beyond Vision-to-Action (VA). In practice, driving VLAs have often trailed VA on planning quality, suggesting that the difficulty is not simply model scale but the interface through which semantic reasoning, temporal context, and continuous control are combined. We argue that this gap reflects how VLA has been built -- as isolated subtask improvements that fail to compose coherent driving capabilities -- rather than what VLA is. We present MindVLA-U1, the first unified streaming VLA architecture for autonomous driving. A unified VLM backbone produces AR language tokens (optional) and flow-matching continuous action trajectories in a single forward pass over one shared representation, preserving the natural output form of each modality. A full streaming design processes the driving video framewise rather than as fixed video-action chunks under costly temporal VLM modeling. Planned trajectories evolve smoothly across frames while a learned streaming memory channel carries temporal context and updates. The unified architecture enables fast/slow systems on dense & sparse MoT backbones via flexible self-attention context management, and exposes a measurable language-control path for action: language-predicted driving intents steers the action diffusion via classifier-free guidance (CFG), turning language-side intent into control signals for continuous action planning. On the long-tail WOD-E2E benchmark, MindVLA-U1 surpasses experienced human drivers for the first time (8.20 RFS vs. 8.13 GT RFS) with 2 diffusion steps, achieves state-of-the-art planning ADEs over prior VA/VLA by large margins, and matches VA latency (16 FPS vs. RAP's 18 FPS at 1B scale) while preserving natural language interfaces for human-vehicle interaction.

While Large Language Models excel in natural language processing, efficiently extending their capabilities to spoken input remains a significant challenge. Existing methods for building SpeechLLMs often rely on computationally expensive full-model fine-tuning, or employ parameter-efficient projectors that suffer from inefficient token sequence lengths and costly full-model supervision. In this paper, we introduce Aligned Continuous Integrate-and-Fire, a highly efficient framework for zero-shot speech processing. Our method dynamically compresses continuous acoustic frames into the exact discrete token length of the target text utilizing explicit Dynamic Time Warping alignments. This allows our initial training stage to establish a robust acoustic-to-semantic bridge using lightweight distance metrics, entirely bypassing the computationally expensive LLM forward pass. For subsequent fine-tuning, we propose a memory-efficient knowledge distillation objective that targets a single LLM layer, performing competitively with full-model cross-entropy training at a fraction of the computational cost. Through extensive evaluations on Automatic Speech Recognition and Speech Translation, we demonstrate that our method achieves superior performance compared to prior parameter-efficient baselines.

Multimodal large language models generate natural-language responses from visual inputs, yet may mention objects absent from an image. In medication assistance, accessible perception, and environmental decision-making, such hallucinations can create real-world safety risks. We propose Semantic-Spatial Agreement Verification (SSAV), a training-free method for verifying object claims. A visually grounded claim should remain stable across semantically equivalent queries and repeatedly localize to the same image region. SSAV aggregates multiple prompts to estimate semantic support and reduce sensitivity to query wording. Query-Induced Regional Verification (QIRV) combines cross-query region persistence, spatial overlap, and relative candidate dominance to identify isolated high responses and dispersed localizations. A geometric mean fuses semantic and spatial evidence, lowering the verification score when either branch lacks support. Experiments on three base models and multiple evaluation protocols show that SSAV effectively mitigates object hallucination. On LLaVA-1.5-7B, accuracy averaged across COCO, A-OKVQA, and GQA improves by 1.81 and 3.17 percentage points under POPE Popular and Adversarial, respectively, while CHAIRs decreases from 49.40% to 32.80%. These results show that cross-query semantic stability and regional consistency provide interpretable external visual evidence for object claims.

While Large Language Model (LLM)-based Natural Language Inference (NLI) systems achieve high accuracy, their decision-making processes lack auditable structures. This paper explores whether NLI can be performed using only interpretable, graph-based representations of evidence. We introduce a fully graph-based pipeline where the classifier never directly processes the input text. Instead, sentences are decomposed into atomic propositions, converted into ConceptNet triples via constrained decoding, and represented as three graphs per pair: premise, hypothesis, and a retrieved ConceptNet subgraph. These graphs are then fed into a fine-tuned 0.8-billion-parameter language model. On the SNLI dataset, our pipeline achieves 89.7% accuracy, just 1.9 points below an identically trained text-based model. On ANLI, it matches the published performance of RoBERTa-large on rounds R2 and R3 (48.0% vs. 48.9% and 44.9% vs. 44.4%) but trails by 16 points on R1, resulting in an overall gap of 9 to 14 points compared to its text counterpart. We term this gap the price of interpretability and demonstrate that it stems from representational limitations rather than data constraints. Ablation studies further reveal that graphs and text are complementary: combining both modalities achieves 92.1% accuracy on SNLI.

LLM-as-a-Judge, which leverages a large language model to evaluate natural language generated by another AI application or model, has become a standard, scalable approach for accelerating and extending costly human evaluation. Yet most work treats a judge as a static artifact, evaluating it once at construction or against a fixed benchmark. We argue instead that an LLM judge operating in a deployed system is better understood as having a lifecycle. It must be built, trained, deployed, and continuously maintained as the surrounding data evolves, and each phase poses distinct technical and operational challenges. We present such a lifecycle for the LLM judges that evaluate recommendation explanations at Netflix. Everything we report comes out of a series of controlled online member-facing experiments, in which our pipeline generated and the judges assessed hundreds of thousands of distinct show-level explanations per week across a changing catalog. Our framework has four phases. (I) Birth defines the evaluation criteria and builds curated benchmark datasets with human labels and rationales. (II) Training refines the judges' rubrics via Reasoning-Aligned Rubric Tuning (RART), which uses a meta-judge over reasoning output as the learning signal. (III) Deployment puts one judge in two online roles, quality gating and reflective generation. (IV) Monitoring runs a continuous Human-in-the-Loop (HITL) alignment process that detects drift and triggers re-tuning behind a human review gate. We report results from a five-week online A/B test over tens of millions of members on the Netflix mobile app, in which judge-aligned explanations shifted member viewing toward novel content (previously unwatched) and increased successful browse-to-play sessions relative to a no-explanation control, with no quality-related escalations.

Multimodal large language models (MLLMs) have demonstrated strong capabilities in vision-language understanding and natural-language response generation. However, these systems can still produce overconfident predictions and hallucination-like outputs, particularly when the visual evidence is weak, ambiguous, or semantically inconsistent. Most existing approaches focus on improving multimodal representation alignment or retrieval-augmented generation, while providing limited mechanisms to quantify instance-level prediction reliability or identify incorrect visual outputs. This work proposes a retrieval-augmented reliability-aware inference framework for trustworthy multimodal visual understanding. The proposed framework constructs an external visual evidence database using pretrained visual embeddings and nearest-neighbor retrieval over normalized feature representations. Retrieved evidence is used to estimate prediction trustworthiness through multiple reliability indicators, including similarity strength, class-support agreement, evidence margin, entropy-based uncertainty, and an aggregate reliability score. Based on these signals, a decision gate determines whether the system should accept the prediction, answer with caution, or abstain/fallback when evidence is insufficient. A multimodal response-generation layer then produces a final user-facing response conditioned on the reliability decision. Experiments on ImageNet-100 demonstrate that the proposed reliability-aware framework improves accepted prediction accuracy from 85.84% to 88.88% at 89.04% coverage. The hallucination-like accepted wrong-answer rate is reduced from 14.16% to 11.12%. These results show that integrating retrieval evidence, reliability estimation, and selective decision gating can improve calibration and reduce overconfident visual errors without retraining large multimodal models.

Recent advances in large language models have enabled natural language to serve as a flexible interface for controlling complex systems, but often at the cost of large-scale multimodal training or weakened domain-specific inductive biases. In structured decision-making domains such as chess, specialized policy networks achieve strong performance but lack semantic controllability, while prompt-conditioned language models are more flexible yet typically exhibit weaker domain grounding. We propose $\textbf{UniMaia}$, a framework for prompt-conditioned policy modulation that adapts a frozen Lc0-based chess policy network using a parameter-efficient text encoder and a ControlNet-style conditioning mechanism. UniMaia enables semantic control over gameplay, including opening selection and player strength, while preserving the pretrained policy representations. We further introduce $\textbf{UniMaia-Aux}$, which incorporates auxiliary temporal conditioning and behavioral prediction objectives. To support this work, we construct a large-scale metadata-augmented Lichess dataset, develop a semi-automated prompt-generation pipeline, and introduce benchmarks spanning both prompt-conditioned and metadata-conditioned settings. UniMaia achieves state-of-the-art expected accuracy on several prompt-conditioned benchmarks and competitive top-move accuracy on general instruction-following tasks, while remaining competitive with dedicated metadata-conditioned approaches on human move prediction benchmarks. UniMaia-Aux further improves expected accuracy and behavioral modeling across several evaluation settings, with modest trade-offs in top-move accuracy. Overall, our results demonstrate that prompt-conditioned control of domain-specific policy networks is feasible without end-to-end multimodal training, while highlighting trade-offs between controllability and predictive performance.

Semantic segmentation across visual modalities such as 3D point clouds and panoramic images remains a challenging task, primarily due to the scarcity of annotated data and the limited adaptability of fixed-label models. In this paper, we present JOPP-3D, an open-vocabulary segmentation framework that jointly leverages panoramic and point cloud data to enable language-driven scene understanding. We convert RGB-D panoramic images into their corresponding wide field-of-view tangential perspectives and 3D point clouds, then use these modalities to extract and align foundational vision-language features. This allows natural language querying to generate semantic masks on both input modalities. Experimental evaluation on the Stanford-2D-3D-s and ToF-360 datasets demonstrates the capability of JOPP-3D to produce coherent and semantically meaningful segmentations across panoramic and 3D domains. Our proposed method achieves a significant improvement compared to the SOTA in open and closed vocabulary 2D and 3D semantic segmentation.

Contemporary studies in mechanistic interpretability have uncovered many puzzling phenomena in the neural information processing of Transformer-based language models, such as induction heads, function vectors, and the Hydra effect. Some of these individual phenomena have been independently tied to different data distributional properties, while some have been loosely associated with model architecture and how Transformers process information. However, a unified understanding of the relationship between data, model architecture, and optimization remains lacking, failing to answer the fundamental question: why do these three phenomena appear universally across different model families and scales, despite their seeming disconnect? In this work, we answer this question by unifying these three phenomena as consequences of hierarchical latent structures in the data generation process, coupled with decorrelated gradients across additive model components and directional concavity in the representation geometry. We validate our theoretical results in a toy model regime and in a large-scale synthetic data regime, comparing them with language models trained on natural language data.

Software increasingly relies on the emergent capabilities of Large Language Models (LLMs), from natural language understanding to program analysis and generation. Yet testing them on specific tasks remains difficult and costly: many prompts lack ground truths, forcing reliance on human judgments, while existing test adequacy measures typically rely on output uncertainty and thus are only available after full inference. A key challenge is to assess how useful a test input is in a way that reflects the demands of the task, ideally before even generating any output. We introduce Clotho, a task-specific, pre-generation test adequacy measure that estimates input difficulty directly from LLM hidden states. Given a large pool of unlabelled inputs for a specific task, Clotho uses a Gaussian Mixture Model (GMM) to adaptively sample the most informative cases for human labelling. Based on this reference set the GMM can then rank unseen inputs by their likelihood of failure. In our empirical evaluation across eight benchmark tasks and three open-weight LLMs, Clotho can predict failures with a ROC-AUC of 0.716, after labelling reference sets that are on average only 5.4% of inputs. It does so without generating any outputs, thereby significantly reducing LLM execution costs compared to output-based uncertainty or confidence measures. Comparison of Clotho and these post-generation adequacy measures shows that the two approaches complement each other. Crucially, we show that adequacy scores learnt from open-weight LLMs transfer effectively to proprietary models, extending the applicability of the approach. When prioritising test inputs for proprietary models, Clotho increases the average number of failing inputs from 18.7 to 42.5 out of 100, compared to random prioritisation.