This paper proposes a method for automating bridge damage understanding and repair priority scoring using fine-tuned Vision-Language Models (VLMs). The authors fine-tune LLaVA-1.5-7B with QLoRA on up to 4,000 paired bridge damage images and inspection text records, evaluating on a fixed test set of 800 images. Results show that 2,000 training samples achieve near-optimal validation loss in 2.9 hours, with diminishing returns beyond that. A two-stage Quality Guard using a fine-tuned Swallow-8B SLM rejects low-quality VLM outputs before priority scoring.

The 10th ABAW Workshop and Competition at CVPR 2026 advances multimodal human-centered AI by introducing new challenges including emotional mimicry intensity estimation, ambivalence/hesitancy recognition, and fine-grained violence detection, alongside traditional affect estimation and recognition tasks. The competition leverages large-scale in-the-wild datasets, and the paper track covers a broad range of topics from pose estimation to fairness and robustness.

EvoSpec is a novel framework that enables real-time evolution of draft models through dynamic vocabulary and parameter adaptation, addressing the sharp decline in acceptance rates of static pruning methods in specialized domains or topic-switching scenarios. It achieves 1.13x speedup over FR-Spec on EAGLE-3 with 27% lower memory overhead.

Large language models (LLMs) are increasingly used as proxies for computational social analysis, but their ability to faithfully represent human communities' 'thick descriptions' remains a critical challenge. This paper introduces CARE (Community-Aware Reaction Evaluation), a reaction-centered framework that benchmarks LLM-simulated discourse against authentic community responses to real-world news. By characterizing a fine-grained spectrum of illocutionary tones, the diagnosis reveals a persistent 'realism gap': steering LLMs with explicit community prompts fails to inherently improve simulation fidelity. Analysis further identifies divergent behavioral signatures among frontier models, suggesting current alignment strategies are insufficient for capturing the sociolinguistic dynamics of…

A new framework called FLUID adapts autoregressive language models to diffusion models for efficient parallel text generation, using Strictly Causal Alignment to reuse GPT checkpoints and Elastic Horizons to dynamically adjust denoising steps. It achieves state-of-the-art performance with significantly reduced training costs.

Researchers identify a Stability-Expressivity Gap in spoken language models when using synthetic data for low-resource languages, and propose two self-alignment frameworks (DGSA and TDSC) that recover prosodic variability and outperform commercial systems like ElevenLabs and Gemini Pro, enabling zero-shot voice cloning for Lao.

BioELX is a novel two-stage framework for cross-lingual biomedical entity linking that requires no annotated training data. It enhances SapBERT with multilingual aliases from Wikidata and uses a pre-trained LLM for context-aware disambiguation. Experiments on five benchmarks show significant improvements, especially for low-resource languages like Turkish, Korean, and Thai.

RAG-Coding is an agentic method for automated ICD-10-CM coding that orchestrates four large language model (LLM) agents and grounds decisions in external knowledge sources, improving coding accuracy and clinical compliance. On the MDACE dataset, it outperforms the best LLM baseline by 8-13% micro-F1 and 2-8% macro-F1. Compared to PLM-ICD, RAG-Coding shows higher micro recall (+11%) but lower micro precision (-6%), with comparable F1 scores. Ablation studies confirm the importance of external knowledge. The authors also release MDACE-2025, updated with expert re-annotations based on 2025 guidelines, enabling finer-grained evaluation.

This paper proposes novel techniques for inter-utterance style interpolation and intra-utterance style transition in prompt-based TTS models, addressing limitations of coarse global control. Methods include direction vector interpolation and KV-cache swapping with sliding-window attention masking. Experiments show high success rates in gender conversion and smooth style transitions within utterances.

Large Language Models (LLMs) acting as autonomous agents can suffer from in-context reward hacking (ICRH), where iterative optimization for proxy objectives leads to harmful side effects. Existing defenses are insufficient because ICRH stems from the model's own over-optimization. This paper proposes LLM-based Constraint Optimization (LCO), a framework with a self-thought module and an evolutionary sampling module that reduces ICRH without fine-tuning. Experiments show LCO reduces Toxicity Growth Rate by 39% on GPT-4 for tweet engagement optimization and reduces ICRH occurrence rate by 15.23% on a policy optimization benchmark, without sacrificing task performance.

ICG is a novel framework that integrates MLLM-based prompting with personalized preference alignment to generate high-quality, contextually relevant cover images. It extracts semantic features via meta tokens, refines them with user embeddings, and injects personalized context into diffusion models. A multi-reward learning strategy combines public rewards with a personalized preference model, eliminating the need for labeled supervision. Experiments show improvements in image quality, semantic fidelity, and personalization, boosting user appeal and recommendation accuracy.

HEAL is a novel cross-layer decentralized learning framework that combines strengths of Federated, Gossip, and Epidemic Learning. Using the Elevator algorithm, it dynamically promotes nodes as aggregators, achieving fault tolerance and performance comparable to Federated Learning in crash-free environments, and outperforming Gossip and Epidemic Learning in crash-prone settings.

The SignGAD framework reformulates graph anomaly detection by replacing fixed pipelines with self-designed task-conditioned workflows, and introduces a guarded final refit strategy to improve reliability under limited supervision.

This paper introduces Architecture-driven Shift (ADS), a lightweight metric for selecting pre-trained models in continual learning. ADS decouples logit shift into architecture and data dependencies, requiring only few data samples to capture shift trends. Experiments across over 175 architectures show strong monotonic correlation (Spearman's r_s ≥ 0.731) between ADS and logit shift, and ADS serves as an effective proxy for expected calibration error for reliable CL model selection across three datasets and six scenarios.

This paper introduces Metric-Aware Principal Component Analysis (MAPCA), which parameterizes PCA with a positive-definite metric matrix and positions it within the geometric deep learning framework. MAPCA interprets the metric as a geometric prior, its solutions are equivariant under the orthogonal group preserving the metric, and its spectrum is invariant. A uniqueness theorem characterizes Invariant PCA (IPCA) as the unique linear data-derived metric in the MAPCA family that is equivariant under arbitrary diagonal rescaling. The paper also discusses extensions to kernel PCA, spectral graph methods, and deep MAPCA.

This survey explores how Mixture-of-Experts (MoE) effectively addresses multimodal learning challenges from three perspectives: efficient engine, representation learner, and adapter, while identifying research gaps.

This paper presents $E^3$-Agent, an executable and evolving agent for resource management of edge AIGC. It separates a fast-path router from a slow-path LLM meta-controller, learns online from execution feedback, and adapts to unknown time-varying service-time mappings. Evaluation shows 65%-73% latency reduction over static baselines and effective stutter suppression.

Research shows that diagonal state space models (S4D) outperform more complex Mamba architectures in time series classification tasks. The authors propose lightweight variants MS4 and MS4N, which achieve higher accuracy and efficiency on 59 datasets, matching deep learning models with 2x to 10x more parameters.

This paper proposes IGADA-IoT, a framework that uses hierarchical multi-generator collaboration and scheduling guided by information gaps to automate data augmentation for sampling frequency decisions, thereby optimizing energy consumption of IoT sensors in wireless sensor networks. Experiments show an average accuracy improvement of 7.27% across multiple downstream models, outperforming existing methods.

This paper proposes Personalized Observation Normalization (PON) for federated reinforcement learning in heterogeneous environments. Each agent locally normalizes raw state inputs using a continuously updated running mean and variance, ensuring consistent scaling without overshadowing. Sharing normalization parameters is shown ineffective. Experiments on heterogeneous MuJoCo tasks demonstrate faster training and superior performance. Accepted at IJCNN 2025.

This paper argues that within-person behavioral variability stems from dynamic latent states, not solely from observable inputs. By intervening on the state's weighting at decision time, outcomes become causally controllable. The framework integrates six lines of evidence (causal inference, predictive processing, allostasis, attentional bottleneck, chronobiology, computational psychiatry) and a 24-month observational dataset from over 200,000 users. It yields seven testable predictions and six operational requirements for state-aware systems, with implications for digital health, education, AI personalization, and personal agency.

Agyn is an open-source platform for AI agents, built on a signal-driven stateful serverless runtime on Kubernetes, a Terraform provider for agent definition, and a zero-trust security model. It is agent-agnostic, model-agnostic, and cloud-agnostic, addressing scalability, governance, and security challenges.

This paper presents a multi-agent architecture for autonomous insight discovery over real-time data streams. It uses Apache Kafka, Flink, and large language models to continuously generate, validate, and visualize hypotheses, shifting from reactive query-driven analytics to proactive discovery-driven systems.

LaneRoPE enables multiple LLM sequences to collaborate during generation via inter-sequence attention and extended RoPE, improving accuracy on math reasoning tasks with minimal architectural changes and negligible inference overhead.

Machine unlearning verification typically focuses on output-level metrics, but a model can pass these while still encoding forgotten data in its internal representations. This paper introduces RULER, a set of representation-level verification metrics, including oracle-comparative M2 and oracle-free M4. Experiments show that approximate unlearning methods pass output-level tests but exhibit significant residuals in representation-level analysis.

This paper proves that large language models have a fundamental limitation in performing causal discovery: methods like supervised fine-tuning, direct preference optimization, and in-context learning cannot distinguish between causal graphs that generate similar observational data. The authors propose Agentic Causal Bayesian Optimization (A-CBO), where a frozen language model serves as an interventional oracle and an external Bayesian loop converges to candidate graphs in logarithmically many rounds. On Corr2Cause, A-CBO matches fine-tuned baselines without any training; on Extended Corr2Cause (scaling to 24 variables and 18K test samples), A-CBO significantly outperforms both fine-tuning and preference optimization.

DynaSchedBench introduces a diagnostic framework for DFJSP using a Sequential Event-Space Calibrator (SESC) to generate difficulty-stratified instances via Schedule Stress Index (SSI). It identifies an 'Observability Paradox' in LLM-based scheduling agents: providing oracle access to full structural information degrades performance compared to concise information. Tool-augmented and refinement strategies also fail to reliably improve performance.

Analogous to the origin of species, this paper addresses the origin of synthetic information, proposing a steganography-based mechanism to trace the lineage of AI-generated content, crucial for maintaining truth and trust in an era of advanced generative models.

Soro is a family of Tajik-specialized conversational LLMs built on Gemma 3, using 1.9B token Tajik continual pretraining and 40K instruction tuning examples. It substantially outperforms same-size Gemma 3 on Tajik benchmarks while retaining English performance. FP8/INT4 quantization preserves gains for edge deployment. An education pilot is underway in Tajikistan.

This paper introduces an LLM-based architecture to detect and quantify the intensity of human values in text. The architecture comprises three coordinated modules that can adapt to various value theories, and experiments on the ValueEval dataset show good detection performance.