We present KokborokMT, a high-quality neural machine translation (NMT) system for Kokborok (ISO 639-3), a Tibeto-Burman language spoken primarily in Tripura, India with approximately 1.5 million speakers. Despite its status as an official language of Tripura, Kokborok has remained severely under-resourced in the NLP community, with prior machine translation attempts limited to systems trained on small Bible-derived corpora achieving BLEU scores below 7. We fine-tune the NLLB-200-distilled-600M model on a multi-source parallel corpus comprising 36,052 sentence pairs: 9,284 professionally translated sentences from the SMOL dataset, 1,769 Bible-domain sentences from WMT shared task data, and 24,999 synthetic back-translated pairs generated via Gemini Flash from Tatoeba English source sentences. We introduce as a new language token for Kokborok in the NLLB framework. Our best system achieves BLEU scores of 17.30 and 38.56 on held-out test sets, representing substantial improvements over prior published results. Human evaluation by three annotators yields mean adequacy of 3.74/5 and fluency of 3.70/5, with substantial agreement between trained evaluators.

Large Language Models (LLMs) exhibit a well-documented positional bias when processing long input contexts: information in the middle of a context window receives substantially less attention than content at the boundaries, a phenomenon termed the Lost-in-the-Middle effect (Liu et al., 2024). This limits knowledge-retrieval applications that embed large structured knowledge bases directly in the LLM context. Retrieval-Augmented Generation (RAG) addresses scalability by retrieving only relevant fragments, but introduces substantial infrastructure overhead and is ill-suited to libraries whose semantic boundaries are human-defined rather than statistically learned. We propose Self-Describing Structured Retrieval (SDSR), a lightweight framework in which structured data files embed human-authored navigational metadata at the file's primacy position, thereby exploiting rather than fighting the LLM's primacy bias. We further propose a Dual-Layer Guidance strategy combining in-file metadata with explicit routing rules in the system prompt. We validate SDSR through a four-round benchmark using a 190-skill library expanded from 36 to 119 categories via adversarial distractor injection. Four conditions are tested: (A) no guidance, (B) in-file summary only, (C) prompt hint only, (D) both combined. Version D achieves 100% primary routing accuracy (20/20) at 119 categories versus 65% for the no-guidance baseline. We identify a fundamental asymmetry: primary routing is solvable by explicit rules, while secondary cross-category routing requires architectural intent explicitly encoded in the data structure. We further extend SDSR to semi-structured corpora, showing how cross-reference encoding enables operation without vector databases in domains with recoverable document structure.

Tuberculosis (TB) is one of the world's deadliest infectious diseases, and in South Africa, it contributes a significant burden to the country's health care system. This paper presents an experimental study on the development of a domain-specific Large Language Model (DS-LLM) for TB care that can help to alleviate the burden on patients and healthcare providers. To achieve this, a literature review was conducted to understand current LLM development strategies, specifically in the medical domain. Thereafter, data were collected from South African TB guidelines, selected TB literature, and existing benchmark medical datasets. We performed LLM fine-tuning by using the Quantised Low-Rank Adaptation (QLoRA) algorithm on a medical LLM (BioMistral-7B), and also implemented Retrieval-Augmented Generation using GraphRAG. The developed DS-LLM was evaluated against the base BioMistral-7B model and a general-purpose LLM using a mix of automated metrics and quantitative ratings. The results show that the DS-LLM had better performance compared to the base model in terms of its contextual alignment (lexical, semantic, and knowledge) for TB care in South Africa.

Large Language Models (LLMs) are increasingly deployed as autonomous agents capable of reasoning, planning, and acting within interactive environments. Despite their growing capability to perform multi-step reasoning and decision-making tasks, internal mechanisms guiding their sequential behavior remain opaque. This paper presents a framework for interpreting the temporal evolution of concepts in LLM agents through a step-wise conformal lens. We introduce the conformal interpretability framework for temporal tasks, which combines step-wise reward modeling with conformal prediction to statistically label model's internal representation at each step as successful or failing. Linear probes are then trained on these representations to identify directions of temporal concepts - latent directions in the model's activation space that correspond to consistent notions of success, failure or reasoning drift. Experimental results on two simulated interactive environments, namely ScienceWorld and AlfWorld, demonstrate that these temporal concepts are linearly separable, revealing interpretable structures aligned with task success. We further show preliminary results on improving an LLM agent's performance by leveraging the proposed framework for steering the identified successful directions inside the model. The proposed approach, thus, offers a principled method for early failure detection as well as intervention in LLM-based agents, paving the path towards trustworthy autonomous language models in complex interactive settings.

Writing discharge summaries to transfer medical information is an important but time-consuming process that can be assisted by Large Language Models (LLMs). This prospective mixed methods pilot study evaluated an Electronic Health Record (EHR)-integrated LLM to generate discharge summaries drafts. In total, 379 discharge summaries were generated in clinical practice by 21 residents and 4 physician assistants during 9 weeks in our academic hospital. LLM-generated text was copied in 58.5% of admissions, and identifiable LLM content could be traced to 29.1% of final discharge letters. Notably, 86.9% of users self-reported a reduction in documentation time, and 60.9% a reduction in administrative workload. Intent to use after the pilot phase was high (91.3%), supporting further implementation of this use-case. Accurately measuring the documentation time of users on discharge summaries remains challenging, but will be necessary for future extrinsic evaluation of LLM-assisted documentation.

The construction of CAD models has traditionally relied on labor-intensive manual operations and specialized expertise. Recent advances in large language models (LLMs) have inspired research into text-to-CAD generation. However, existing approaches typically treat generation and editing as disjoint tasks, limiting their practicality. We propose PR-CAD, a progressive refinement framework that unifies generation and editing for controllable and faithful text-to-CAD modeling. To support this, we curate a high-fidelity interaction dataset spanning the full CAD lifecycle, encompassing multiple CAD representations as well as both qualitative and quantitative descriptions. The dataset systematically defines the types of edit operations and generates highly human-like interaction data. Building on a CAD representation tailored for LLMs, we propose a reinforcement learning-enhanced reasoning framework that integrates intent understanding, parameter estimation, and precise edit localization into a single agent. This enables an "all-in-one" solution for both design creation and refinement. Extensive experiments demonstrate strong mutual reinforcement between generation and editing tasks, and across qualitative and quantitative modalities. On public benchmarks, PR-CAD achieves state-of-the-art controllability and faithfulness in both generation and refinement scenarios, while also proving user-friendly and significantly improving CAD modeling efficiency.

Large language models (LLMs) are increasingly used in scientific writing but struggle with book-length tasks, often producing inconsistent structure and unreliable citations. We introduce CoAuthorAI, a human-in-the-loop writing system that combines retrieval-augmented generation, expert-designed hierarchical outlines, and automatic reference linking. The system allows experts to iteratively refine text at the sentence level, ensuring coherence and accuracy. In evaluations of 500 multi-domain literature review chapters, CoAuthorAI achieved a maximum soft-heading recall of 98%; in a human evaluation of 100 articles, the generated content reached a satisfaction rate of 82%. The book AI for Rock Dynamics generated with CoAuthorAI and Kexin Technology's LUFFA AI model has been published with Springer Nature. These results show that systematic human-AI collaboration can extend LLMs' capabilities from articles to full-length books, enabling faster and more reliable scientific publishing.

LLM agents lack persistent memory, causing conversations to reset each session and preventing personalization over time. We present Lyzr Cognis, a unified memory architecture for conversational AI agents that addresses this limitation through a multi-stage retrieval pipeline. Cognis combines a dual-store backend pairing OpenSearch BM25 keyword matching with Matryoshka vector similarity search, fused via Reciprocal Rank Fusion. Its context-aware ingestion pipeline retrieves existing memories before extraction, enabling intelligent version tracking that preserves full memory history while keeping the store consistent. Temporal boosting enhances time-sensitive queries, and a BGE-2 cross-encoder reranker refines final result quality. We evaluate Cognis on two independent benchmarks -- LoCoMo and LongMemEval -- across eight answer generation models, demonstrating state-of-the-art performance on both. The system is open-source and deployed in production serving conversational AI applications.

We present a hybrid multi-phase page matching algorithm for automated comparison of Japanese building permit document sets. Building permit review in Japan requires cross-referencing large PDF document sets across revision cycles, a process that is labor-intensive and error-prone when performed manually. The algorithm combines longest common subsequence (LCS) structural alignment, a seven-phase consensus matching pipeline, and a dynamic programming optimal alignment stage to robustly pair pages across revisions even when page order, numbering, or content changes substantially. A subsequent multi-layer diff engine -- comprising text-level, table-level, and pixel-level visual differencing -- produces highlighted difference reports. Evaluation on real-world permit document sets achieves F1=0.80 and precision=1.00 on a manually annotated ground-truth benchmark, with zero false-positive matched pairs.

Key-value (KV) caching is critical for efficient inference in large language models (LLMs), yet its memory footprint scales linearly with context length, resulting in a severe scalability bottleneck. Existing approaches largely treat KV states as equally important across time, implicitly assuming uniform precision and accessibility. However, this assumption contrasts with human memory systems, where memories vary in clarity, recall frequency, and relevance with temporal proximity.Motivated by this insight, we propose TTKV, a KV cache management framework that maps the human memory system onto the KV cache. TTKV partitions the KV cache into temporal tiers with heterogeneous capacity and precision. The design addresses three aspects: (1) Tier Layout, decoupling fast and slow memory using HBM and DRAM; (2) Tier Content, assigning more recent KV states to faster, higher-precision tiers based on temporal proximity; and (3) Tier Interaction, employing block-wise streaming attention to overlap communication and computation when accessing slow tiers. Experiments show that TTKV reduces cross-tier traffic by 5.94x on 128K-context tasks, achieving up to 76% latency reduction and 2x throughput improvement over strong baselines.

Large language models (LLMs) exhibit systematic miscalibration with rhetorical intensity not proportionate to epistemic grounding. This study tests this hypothesis and proposes a framework for quantifying this decoupling by designing a triadic epistemic-rhetorical marker (ERM) taxonomy. The taxonomy is operationalized through composite metrics of form-meaning divergence (FMD), genuine-to-performed epistemic ratio (GPR), and rhetorical device distribution entropy (RDDE). Applied to 225 argumentative texts spanning approximately 0.6 Million tokens across human expert, human non-expert, and LLM-generated sub-corpora, the framework identifies a consistent, model-agnostic LLM epistemic signature. LLM-generated texts produce tricolon at nearly twice the expert rate ($Δ= 0.95$), while human authors produce erotema at more than twice the LLM rate. Performed hesitancy markers appear at twice the human density in LLM output. FMD is significantly elevated in LLM texts relative to both human groups ($p < 0.001, Δ= 0.68$), and rhetorical devices are distributed significantly more uniformly across LLM documents. The findings are consistent with theoretical intuitions derived from Gricean pragmatics, Relevance Theory, and Brandomian inferentialism. The annotation pipeline is fully automatable, making it deployable as a lightweight screening tool for epistemic miscalibration in AI-generated content and as a theoretically motivated feature set for LLM-generated text detection pipelines.

We evaluate speculative decoding with EAGLE3 as an inference-time optimization for PayPal's Commerce Agent, powered by a fine-tuned llama3.1-nemotron-nano-8B-v1 model. Building on prior work (NEMO-4-PAYPAL) that reduced latency and cost through domain-specific fine-tuning, we benchmark EAGLE3 via vLLM against NVIDIA NIM on identical 2xH100 hardware across 40 configurations spanning speculative token counts (gamma=3, gamma=5), concurrency levels (1-32), and sampling temperatures (0, 0.5). Key findings: (1) gamma=3 achieves 22-49% throughput improvement and 18-33% latency reduction at zero additional hardware cost; (2) acceptance rates remain stable at approximately 35.5% for gamma=3 across all conditions; (3) gamma=5 yields diminishing returns (approximately 25% acceptance rate); (4) LLM-as-Judge evaluation confirms fully preserved output quality; and (5) speculative decoding on a single H100 matches or exceeds NIM on two H100s, enabling 50% GPU cost reduction.

Retrieval-Augmented Generation (RAG) expands the knowledge of Large Language Models (LLMs), yet current static retrieval methods struggle with complex, multi-hop problems. While recent dynamic retrieval strategies offer improvements, they face two key challenges: 1) irrelevant retrieved noise can misdirect the reasoning process, and 2) processing full documents incurs prohibitive computational and latency costs. To address these issues, we propose OThink-SRR1, a framework that enhances large models with an iterative Search-Refine-Reason process trained via reinforcement learning. Its core Refine stage distills retrieved documents into concise, relevant facts before reasoning. We introduce GRPO-IR, an end-to-end reinforcement learning algorithm that rewards accurate evidence identification while penalizing excessive retrievals, thus training the model to be both focused and efficient. Experiments on four multi-hop QA benchmarks show our approach achieves superior accuracy over strong baselines while using fewer retrieval steps and tokens. This positions OThink-SRR1 as a potent foundational model for information-seeking agents.

Recent work identifies a sparse set of "hallucination neurons" (H-neurons), less than 0.1% of feed-forward network neurons, that reliably predict when large language models will hallucinate. These neurons are identified on general-knowledge question answering and shown to generalize to new evaluation instances. We ask a natural follow-up question: do H-neurons generalize across knowledge domains? Using a systematic cross-domain transfer protocol across 6 domains (general QA, legal, financial, science, moral reasoning, and code vulnerability) and 5 open-weight models (3B to 8B parameters), we find they do not. Classifiers trained on one domain's H-neurons achieve AUROC 0.783 within-domain but only 0.563 when transferred to a different domain (delta = 0.220, p < 0.001), a degradation consistent across all models tested. Our results suggest that hallucination is not a single mechanism with a universal neural signature, but rather involves domain-specific neuron populations that differ depending on the knowledge type being queried. This finding has direct implications for the deployment of neuron-level hallucination detectors, which must be calibrated per domain rather than trained once and applied universally.

Stereotypes in large language models (LLMs) can perpetuate harmful societal biases. Despite the widespread use of models, little is known about where these biases reside in the neural network. This study investigates the internal mechanisms of GPT 2 Small and Llama 3.2 to locate stereotype related activations. We explore two approaches: identifying individual contrastive neuron activations that encode stereotypes, and detecting attention heads that contribute heavily to biased outputs. Our experiments aim to map these "bias fingerprints" and provide initial insights for mitigating stereotypes.

Modern machine learning is still largely organized around a single recipe: choose a parameterized model family and optimize its weights. Although highly successful, this paradigm is too narrow for many structured prediction problems, where the main bottleneck is not parameter fitting but discovering what should be computed from the data. Success often depends on identifying the right transformations, statistics, invariances, interaction structures, temporal summaries, gates, or nonlinear compositions, especially when objectives are non-differentiable, evaluation is cross-validation-based, interpretability matters, or continual adaptation is required. We present EvoForest, a hybrid neuro-symbolic system for end-to-end open-ended evolution of computation. Rather than merely generating features, EvoForest jointly evolves reusable computational structure, callable function families, and trainable low-dimensional continuous components inside a shared directed acyclic graph. Intermediate nodes store alternative implementations, callable nodes encode reusable transformation families such as projections, gates, and activations, output nodes define candidate predictive computations, and persistent global parameters can be refined by gradient descent. For each graph configuration, EvoForest evaluates the discovered computation and uses a lightweight Ridge-based readout to score the resulting representation against a non-differentiable cross-validation target. The evaluator also produces structured feedback that guides future LLM-driven mutations. In the 2025 ADIA Lab Structural Break Challenge, EvoForest reached 94.13% ROC-AUC after 600 evolution steps, exceeding the publicly reported winning score of 90.14% under the same evaluation protocol.

We present a simulation-based evaluation of the Inference Headroom Ratio (IHR), a dimensionless diagnostic quantity for characterizing inference stability in constrained decision systems. IHR formalizes the relationship between a system's effective inferential capacity C and the combined uncertainty and constraint load U + K imposed by its operating environment, and is intended to capture proximity to an inference stability boundary rather than output-level performance. Across three controlled experiments, we show that IHR functions as: (1) a quantifiable risk indicator whose relationship to collapse probability follows a well-fitted logistic curve with estimated critical threshold IHR* approx. 1.19, (2) a sensitive indicator of proximity to the inference stability boundary under environmental noise, and (3) a viable control variable whose active regulation reduces system collapse rate from 79.4% to 58.7% and IHR variance by 70.4% across 300 Monte Carlo runs. These results position IHR as a prospective, system-level complement to standard performance, drift, and uncertainty metrics, enabling estimation of remaining inferential margin before overt failure in AI systems operating under distributional shift and constraint.

Clinical trials require strict adherence to medication protocols, yet dosing errors remain a persistent challenge affecting patient safety and trial integrity. We present an automated system for detecting dosing errors in unstructured clinical trial narratives using gradient boosting with comprehensive multi-modal feature engineering. Our approach combines 3,451 features spanning traditional NLP (TF-IDF, character n-grams), dense semantic embeddings (all-MiniLM-L6v2), domain-specific medical patterns, and transformer-based scores (BiomedBERT, DeBERTa-v3), used to train a LightGBM model. Features are extracted from nine complementary text fields (median 5,400 characters per sample) ensuring complete coverage across all 42,112 clinical trial narratives. On the CT-DEB benchmark dataset with severe class imbalance (4.9% positive rate), we achieve 0.8725 test ROC-AUC through 5-fold ensemble averaging (cross-validation: 0.8833 + 0.0091 AUC). Systematic ablation studies reveal that removing sentence embeddings causes the largest performance degradation (2.39%), demonstrating their critical role despite contributing only 37.07% of total feature importance. Feature efficiency analysis demonstrates that selecting the top 500-1000 features yields optimal performance (0.886-0.887 AUC), outperforming the full 3,451-feature set (0.879 AUC) through effective noise reduction. Our findings highlight the importance of feature selection as a regularization technique and demonstrate that sparse lexical features remain complementary to dense representations for specialized clinical text classification under severe class imbalance.

We present ThermoQA, a benchmark of 293 open-ended engineering thermodynamics problems in three tiers: property lookups (110 Q), component analysis (101 Q), and full cycle analysis (82 Q). Ground truth is computed programmatically from CoolProp 7.2.0, covering water, R-134a, and variable-cp air. Six frontier LLMs are evaluated across three independent runs each. The composite leaderboard is led by Claude Opus 4.6 (94.1%), GPT-5.4 (93.1%), and Gemini 3.1 Pro (92.5%). Cross-tier degradation ranges from 2.8 pp (Opus) to 32.5 pp (MiniMax), confirming that property memorization does not imply thermodynamic reasoning. Supercritical water, R-134a refrigerant, and combined-cycle gas turbine analysis serve as natural discriminators with 40-60 pp performance spreads. Multi-run sigma ranges from +/-0.1% to +/-2.5%, quantifying reasoning consistency as a distinct evaluation axis. Dataset and code are open-source at https://huggingface.co/datasets/olivenet/thermoqa

This paper presents a transparent screening framework for estimating inference and training impacts of current large language models under limited observability. The framework converts natural-language application descriptions into bounded environmental estimates and supports a comparative online observatory of current market models. Rather than claiming direct measurement for opaque proprietary services, it provides an auditable, source-linked proxy methodology designed to improve comparability, transparency, and reproducibility.

Large language model (LLM) agents often suffer from high reasoning overhead, excessive token consumption, unstable execution, and inability to reuse past experiences in complex tasks like business queries, tool use, and workflow orchestration. Traditional methods generate workflows from scratch for every query, leading to high cost, slow response, and poor robustness. We propose WorkflowGen, an adaptive, trajectory experience-driven framework for automatic workflow generation that reduces token usage and improves efficiency and success rate. Early in execution, WorkflowGen captures full trajectories and extracts reusable knowledge at both node and workflow levels, including error fingerprints, optimal tool mappings, parameter schemas, execution paths, and exception-avoidance strategies. It then employs a closed-loop mechanism that performs lightweight generation only on variable nodes via trajectory rewriting, experience updating, and template induction. A three-tier adaptive routing strategy dynamically selects among direct reuse, rewriting-based generation, and full initialization based on semantic similarity to historical queries. Without large annotated datasets, we qualitatively compare WorkflowGen against real-time planning, static single trajectory, and basic in-context learning baselines. Our method reduces token consumption by over 40 percent compared to real-time planning, improves success rate by 20 percent on medium-similarity queries through proactive error avoidance and adaptive fallback, and enhances deployability via modular, traceable experiences and cross-scenario adaptability. WorkflowGen achieves a practical balance of efficiency, robustness, and interpretability, addressing key limitations of existing approaches.

Automated scoring of students' scientific explanations offers the potential for immediate, accurate feedback, yet class imbalance in rubric categories particularly those capturing advanced reasoning remains a challenge. This study investigates augmentation strategies to improve transformer-based text classification of student responses to a physical science assessment based on an NGSS-aligned learning progression. The dataset consists of 1,466 high school responses scored on 11 binary-coded analytic categories. This rubric identifies six important components including scientific ideas needed for a complete explanation along with five common incomplete or inaccurate ideas. Using SciBERT as a baseline, we applied fine-tuning and test these augmentation strategies: (1) GPT-4--generated synthetic responses, (2) EASE, a word-level extraction and filtering approach, and (3) ALP (Augmentation using Lexicalized Probabilistic context-free grammar) phrase-level extraction. While fine-tuning SciBERT improved recall over baseline, augmentation substantially enhanced performance, with GPT data boosting both precision and recall, and ALP achieving perfect precision, recall, and F1 scores across most severe imbalanced categories (5,6,7 and 9). Across all rubric categories EASE augmentation substantially increased alignment with human scoring for both scientific ideas (Categories 1--6) and inaccurate ideas (Categories 7--11). We compared different augmentation strategies to a traditional oversampling method (SMOTE) in an effort to avoid overfitting and retain novice-level data critical for learning progression alignment. Findings demonstrate that targeted augmentation can address severe imbalance while preserving conceptual coverage, offering a scalable solution for automated learning progression-aligned scoring in science education.

We propose a feature-free approach to algorithm selection that replaces hand-crafted instance features with pretrained text embeddings. Our method, ZeroFolio, proceeds in three steps: it reads the raw instance file as plain text, embeds it with a pretrained embedding model, and selects an algorithm via weighted k-nearest neighbors. The key to our approach is the observation that pretrained embeddings produce representations that distinguish problem instances without any domain knowledge or task-specific training. This allows us to apply the same three-step pipeline (serialize, embed, select) across diverse problem domains with text-based instance formats. We evaluate our approach on 11 ASlib scenarios spanning 7 domains (SAT, MaxSAT, QBF, ASP, CSP, MIP, and graph problems). Our experiments show that this approach outperforms a random forest trained on hand-crafted features in 10 of 11 scenarios with a single fixed configuration, and in all 11 with two-seed voting; the margin is often substantial. Our ablation study shows that inverse-distance weighting, line shuffling, and Manhattan distance are the key design choices. On scenarios where both selectors are competitive, combining embeddings with hand-crafted features via soft voting yields further improvements.

Generative AI is entering research, education, and professional work faster than current governance frameworks can specify how AI-assisted outputs should be judged in learning-intensive settings. The central problem is proxy failure: a polished artifact can be useful while no longer serving as credible evidence of the human understanding, judgment, or transfer ability that the work is supposed to cultivate or certify. This paper proposes AI to Learn 2.0, a deliverable-oriented governance framework for AI-assisted work. Rather than claiming element-wise novelty, it reorganizes adjacent ideas around the final deliverable package, distinguishes artifact residual from capability residual, and operationalizes the result through a five-part package, a seven-dimension maturity rubric, gate thresholds on critical dimensions, and a companion capability-evidence ladder. AI to Learn 2.0 allows opaque AI during exploration, drafting, hypothesis generation, and workflow design, but requires that the released deliverable be usable, auditable, transferable, and justifiable without the original large language model or cloud API. In learning-intensive contexts, it additionally requires context-appropriate human-attributable evidence of explanation or transfer. Worked scoring across contrastive cases, including coursework substitution, a symbolic-regression governance contrast, teacher-audited national-exam practice forms, and a self-hosted lecture-to-quiz pipeline with deterministic quality control, shows how the framework separates polished substitution workflows from bounded, auditable, and handoff-ready AI-assisted workflows. AI to Learn 2.0 is proposed as a governance instrument for structured third-party review where capability preservation, accountability, and validity boundaries matter.

Equipping LLMs with external tools effectively addresses internal reasoning limitations. However, it introduces a critical yet under-explored phenomenon: tool overuse, the unnecessary tool-use during reasoning. In this paper, we first reveal this phenomenon is pervasive across diverse LLMs. We then experimentally elucidate its underlying mechanisms through two key lenses: (1) First, by analyzing tool-use behavior across different internal knowledge availability regions, we identify a \textit{knowledge epistemic illusion}: models misjudge internal knowledge boundaries and fail to accurately perceive their actual knowledge availability. To mitigate this, we propose a knowledge-aware epistemic boundary alignment strategy based on direct preference optimization, which reduces tool usage in by 82.8\% while yielding an accuracy improvement. (2) Second, we establish a causal link between reward structures and tool-use behavior by visualizing the tool-augmented training process. It reveals that \textit{outcome-only rewards} inadvertently encourage tool overuse by rewarding only final correctness, regardless of tool efficiency. To verify this, we balance reward signals during training rather than relying on outcome-only rewards, cutting unnecessary tool calls by 66.7\% (7B) and 60.7\% (32B) without sacrificing accuracy. Finally, we provide theoretical justification in this two lenses to understand tool overuse.

Recent advances in Diffusion Transformers (DiTs) have enabled high-quality joint audio-video generation, producing videos with synchronized audio within a single model. However, existing controllable generation frameworks are typically restricted to video-only control. This restricts comprehensive controllability and often leads to suboptimal cross-modal alignment. To bridge this gap, we present MMControl, which enables users to perform Multi-Modal Control in joint audio-video generation. MMControl introduces a dual-stream conditional injection mechanism. It incorporates both visual and acoustic control signals, including reference images, reference audio, depth maps, and pose sequences, into a joint generation process. These conditions are injected through bypass branches into a joint audio-video Diffusion Transformer, enabling the model to simultaneously generate identity-consistent video and timbre-consistent audio under structural constraints. Furthermore, we introduce modality-specific guidance scaling, which allows users to independently and dynamically adjust the influence strength of each visual and acoustic condition at inference time. Extensive experiments demonstrate that MMControl achieves fine-grained, composable control over character identity, voice timbre, body pose, and scene layout in joint audio-video generation.

Fine-grained high-resolution remote sensing mapping typically relies on localized visual features, which restricts cross-domain generalizability and often leads to fragmented predictions of large-scale land covers. While global geospatial foundation models offer powerful, generalizable representations, directly fusing their high-dimensional implicit embeddings with high-resolution visual features frequently triggers feature interference and spatial structure degradation due to a severe semantic-spatial gap. To overcome these limitations, we propose a Structure-Semantic Decoupled Modulation (SSDM) framework, which decouples global geospatial representations into two complementary cross-modal injection pathways. First, the structural prior modulation branch introduces the macroscopic receptive field priors from global representations into the self-attention modules of the high-resolution encoder. By guiding local feature extraction with holistic structural constraints, it effectively suppresses prediction fragmentation caused by high-frequency detail noise and excessive intra-class variance. Second, the global semantic injection branch explicitly aligns holistic context with the deep high-resolution feature space and directly supplements global semantics via cross-modal integration, thereby significantly enhancing the semantic consistency and category-level discrimination of complex land covers. Extensive experiments demonstrate that our method achieves state-of-the-art performance compared to existing cross-modal fusion approaches. By unleashing the potential of global embeddings, SSDM consistently improves high-resolution mapping accuracy across diverse scenarios, providing a universal and effective paradigm for integrating geospatial foundation models into high-resolution vision tasks.

Egocentric assistants often rely on first-person view data to capture user behavior and context for personalized services. Since different users exhibit distinct habits, preferences, and routines, such personalization is essential for truly effective assistance. However, effectively integrating long-term user data for personalization remains a key challenge. To address this, we introduce EgoSelf, a system that includes a graph-based interaction memory constructed from past observations and a dedicated learning task for personalization. The memory captures temporal and semantic relationships among interaction events and entities, from which user-specific profiles are derived. The personalized learning task is formulated as a prediction problem where the model predicts possible future interactions from individual user's historical behavior recorded in the graph. Extensive experiments demonstrate the effectiveness of EgoSelf as a personalized egocentric assistant. Code is available at https://abie-e.github.io/EgoSelf/.

The rapid adoption of Large Language Models (LLMs) has spurred interest in automated peer review; however, progress is currently stifled by benchmarks that treat reviewing primarily as a rating prediction task. We argue that the utility of a review lies in its textual justification--its arguments, questions, and critique--rather than a scalar score. To address this, we introduce Beyond Rating, a holistic evaluation framework that assesses AI reviewers across five dimensions: Content Faithfulness, Argumentative Alignment, Focus Consistency, Question Constructiveness, and AI-Likelihood. Notably, we propose a Max-Recall strategy to accommodate valid expert disagreement and introduce a curated dataset of paper with high-confidence reviews, rigorously filtered to remove procedural noise. Extensive experiments demonstrate that while traditional n-gram metrics fail to reflect human preferences, our proposed text-centric metrics--particularly the recall of weakness arguments--correlate strongly with rating accuracy. These findings establish that aligning AI critique focus with human experts is a prerequisite for reliable automated scoring, offering a robust standard for future research.

Composed Image Retrieval (CIR) has attracted significant attention due to its flexible multimodal query method, yet its development is severely constrained by the Noisy Triplet Correspondence (NTC) problem. Most existing robust learning methods rely on the "small loss hypothesis", but the unique semantic ambiguity in NTC, such as "partial matching", invalidates this assumption, leading to unreliable noise identification. This entraps the model in a self dependent vicious cycle where the learner is intertwined with the arbiter, ultimately causing catastrophic "representation pollution". To address this critical challenge, we propose a novel "Expert-Proxy-Diversion" decoupling paradigm, named Air-Know (ArbIteR calibrated Knowledge iNternalizing rObust netWork). Air-Know incorporates three core modules: (1) External Prior Arbitration (EPA), which utilizes Multimodal Large Language Models (MLLMs) as an offline expert to construct a high precision anchor dataset; (2) Expert Knowledge Internalization (EKI), which efficiently guides a lightweight proxy "arbiter" to internalize the expert's discriminative logic; (3) Dual Stream Reconciliation (DSR), which leverages the EKI's matching confidence to divert the training data, achieving a clean alignment stream and a representation feedback reconciliation stream. Extensive experiments on multiple CIR benchmark datasets demonstrate that Air-Know significantly outperforms existing SOTA methods under the NTC setting, while also showing strong competitiveness in traditional CIR.