Humans vs Vision-Language Models: A Unified Measure of Narrative Coherence

We study narrative coherence in visually grounded stories by comparing human-written narratives with those generated by vision-language models (VLMs) on the Visual Writing Prompts corpus. Using a set of metrics that capture different aspects of narrative coherence, including coreference, discourse relation types, topic continuity, character persistence, and multimodal character grounding, we compute a narrative coherence score. We find that VLMs show broadly similar coherence profiles that differ systematically from those of humans. In addition, differences for individual measures are often subtle, but they become clearer when considered jointly. Overall, our results indicate that, despite human-like surface fluency, model narratives exhibit systematic differences from those of humans in how they organise discourse across a visually grounded story. Our code is available at https://github.com/GU-CLASP/coherence-driven-humans.

SozKZ: Training Efficient Small Language Models for Kazakh from Scratch

Kazakh, a Turkic language spoken by over 22 million people, remains underserved by existing multilingual language models, which allocate minimal capacity to low-resource languages and employ tokenizers ill-suited to agglutinative morphology. We present SozKZ, a family of Llama-architecture language models (50M-600M parameters) trained entirely from scratch on 9 billion tokens of Kazakh text with a dedicated 50K BPE tokenizer. We evaluate all models on three Kazakh benchmarks -- multiple-choice cultural QA, reading comprehension (Belebele), and topic classification (SIB-200) -- alongside five multilingual baselines ranging from 500M to 3B parameters. Our 600M model achieves 30.3% accuracy on Kazakh cultural QA, approaching the 32.0% of Llama-3.2-1B (2x larger), and 25.5% on SIB-200 topic classification, surpassing all evaluated multilingual models up to 2B parameters. We observe consistent scaling from 50M to 600M, with MC QA accuracy rising from 22.8% to 30.3%, suggesting that further scaling remains beneficial. These results demonstrate that small, dedicated models trained from scratch with a language-appropriate tokenizer offer a viable path for low-resource language technology, achieving competitive performance at a fraction of the computational cost. All models and the tokenizer are released under open licenses.

When AI Shows Its Work, Is It Actually Working? Step-Level Evaluation Reveals Frontier Language Models Frequently Bypass Their Own Reasoning

Language models increasingly "show their work" by writing step-by-step reasoning before answering. But are these reasoning steps genuinely used, or decorative narratives generated after the model has already decided? Consider: a medical AI writes "The patient's eosinophilia and livedo reticularis following catheterization suggest cholesterol embolization syndrome. Answer: B." If we remove the eosinophilia observation, does the diagnosis change? For most frontier models, the answer is no - the step was decorative. We introduce step-level evaluation: remove one reasoning sentence at a time and check whether the answer changes. This simple test requires only API access -- no model weights -- and costs approximately $1-2 per model per task. Testing 10 frontier models (GPT-5.4, Claude Opus, DeepSeek-V3.2, MiniMax-M2.5, Kimi-K2.5, and others) across sentiment, mathematics, topic classification, and medical QA (N=376-500 each), the majority produce decorative reasoning: removing any step changes the answer less than 17% of the time, while any single step alone recovers the answer. This holds even on math, where smaller models (0.8-8B) show genuine step dependence (55% necessity). Two models break the pattern: MiniMax-M2.5 on sentiment (37% necessity) and Kimi-K2.5 on topic classification (39%) - but both shortcut other tasks. Faithfulness is model-specific and task-specific. We also discover "output rigidity": on the same medical questions, Claude Opus writes 11 diagnostic steps while GPT-OSS-120B outputs a single token. Mechanistic analysis (attention patterns) confirms that CoT attention drops more in late layers for decorative tasks (33%) than faithful ones (20%). Implications: step-by-step explanations from frontier models are largely decorative, per-model per-domain evaluation is essential, and training objectives - not scale - determine whether reasoning is genuine.

ByteRover: Agent-Native Memory Through LLM-Curated Hierarchical Context

Memory-Augmented Generation (MAG) extends large language models with external memory to support long-context reasoning, but existing approaches universally treat memory as an external service that agents call into, delegating storage to separate pipelines of chunking, embedding, and graph extraction. This architectural separation means the system that stores knowledge does not understand it, leading to semantic drift between what the agent intended to remember and what the pipeline actually captured, loss of coordination context across agents, and fragile recovery after failures. In this paper, we propose ByteRover, an agent-native memory architecture that inverts the memory pipeline: the same LLM that reasons about a task also curates, structures, and retrieves knowledge. ByteRover represents knowledge in a hierarchical Context Tree, a file-based knowledge graph organized as Domain, Topic, Subtopic, and Entry, where each entry carries explicit relations, provenance, and an Adaptive Knowledge Lifecycle (AKL) with importance scoring, maturity tiers, and recency decay. Retrieval uses a 5-tier progressive strategy that resolves most queries at sub-100 ms latency without LLM calls, escalating to agentic reasoning only for novel questions. Experiments on LoCoMo and LongMemEval demonstrate that ByteRover achieves state-of-the-art accuracy on LoCoMo and competitive results on LongMemEval while requiring zero external infrastructure, no vector database, no graph database, no embedding service, with all knowledge stored as human-readable markdown files on the local filesystem.

ImagenWorld: Stress-Testing Image Generation Models with Explainable Human Evaluation on Open-ended Real-World Tasks

Advances in diffusion, autoregressive, and hybrid models have enabled high-quality image synthesis for tasks such as text-to-image, editing, and reference-guided composition. Yet, existing benchmarks remain limited, either focus on isolated tasks, cover only narrow domains, or provide opaque scores without explaining failure modes. We introduce \textbf{ImagenWorld}, a benchmark of 3.6K condition sets spanning six core tasks (generation and editing, with single or multiple references) and six topical domains (artworks, photorealistic images, information graphics, textual graphics, computer graphics, and screenshots). The benchmark is supported by 20K fine-grained human annotations and an explainable evaluation schema that tags localized object-level and segment-level errors, complementing automated VLM-based metrics. Our large-scale evaluation of 14 models yields several insights: (1) models typically struggle more in editing tasks than in generation tasks, especially in local edits. (2) models excel in artistic and photorealistic settings but struggle with symbolic and text-heavy domains such as screenshots and information graphics. (3) closed-source systems lead overall, while targeted data curation (e.g., Qwen-Image) narrows the gap in text-heavy cases. (4) modern VLM-based metrics achieve Kendall accuracies up to 0.79, approximating human ranking, but fall short of fine-grained, explainable error attribution. ImagenWorld provides both a rigorous benchmark and a diagnostic tool to advance robust image generation.

Dual-branch Graph Domain Adaptation for Cross-scenario Multi-modal Emotion Recognition

Multimodal Emotion Recognition in Conversations (MERC) aims to predict speakers' emotional states in multi-turn dialogues through text, audio, and visual cues. In real-world settings, conversation scenarios differ significantly in speakers, topics, styles, and noise levels. Existing MERC methods generally neglect these cross-scenario variations, limiting their ability to transfer models trained on a source domain to unseen target domains. To address this issue, we propose a Dual-branch Graph Domain Adaptation framework (DGDA) for multimodal emotion recognition under cross-scenario conditions. We first construct an emotion interaction graph to characterize complex emotional dependencies among utterances. A dual-branch encoder, consisting of a hypergraph neural network (HGNN) and a path neural network (PathNN), is then designed to explicitly model multivariate relationships and implicitly capture global dependencies. To enable out-of-domain generalization, a domain adversarial discriminator is introduced to learn invariant representations across domains. Furthermore, a regularization loss is incorporated to suppress the negative influence of noisy labels. To the best of our knowledge, DGDA is the first MERC framework that jointly addresses domain shift and label noise. Theoretical analysis provides tighter generalization bounds, and extensive experiments on IEMOCAP and MELD demonstrate that DGDA consistently outperforms strong baselines and better adapts to cross-scenario conversations. Our code is available at https://github.com/Xudmm1239439/DGDA-Net.

Indian Peak Power demand Forecasting : Transformer Based Implementation of Temporal Architecture

The long-term forecasting of electricity demand has been a prevalent research topic, primarily because of its economic and strategic relevance. Several machine learning as well as deep learning techniques have been developed in parallel with the growing complexity of the peak demand, planning for generation facilities and transmission augmentation in future. Most of these proposed techniques work on short-term forecasting as long-term forecasting is considerably more challenging due to unpredictable and unforeseeable variables that may arise in the future. This paper proposes a Temporal Fusion Transformer based deep learning approach for long term forecasting of peak power demand. The dataset used in this paper consists of peak power demand in India for a period of 6 years and the prediction was done for a period of 1 year. Our proposed model was compared with other popular forecasting models and it performed considerably better in benchmarks and was also more accurate in modelling the variance in the power demand.

VAANI: Capturing the language landscape for an inclusive digital India

Project VAANI is an initiative to create an India-representative multi-modal dataset that comprehensively maps India's linguistic diversity, starting with 165 districts across the country in its first two phases. Speech data is collected through a carefully structured process that uses image-based prompts to encourage spontaneous responses. Images are captured through a separate process that encompasses a broad range of topics, gathered from both within and across districts. The collected data undergoes a rigorous multi-stage quality evaluation, including both automated and manual checks to ensure highest possible standards in audio quality and transcription accuracy. Following this thorough validation, we have open-sourced around 289K images, approximately 31,270 hours of audio recordings, and around 2,067 hours of transcribed speech, encompassing 112 languages from 165 districts from 31 States and Union territories. Notably, significant of these languages are being represented for the first time in a dataset of this scale, making the VAANI project a groundbreaking effort in preserving and promoting linguistic inclusivity. This data can be instrumental in building inclusive speech models for India, and in advancing research and development across speech, image, and multimodal applications.

From Content to Audience: A Multimodal Annotation Framework for Broadcast Television Analytics

Automated semantic annotation of broadcast television content presents distinctive challenges, combining structured audiovisual composition, domain-specific editorial patterns, and strict operational constraints. While multimodal large language models (MLLMs) have demonstrated strong general-purpose video understanding capabilities, their comparative effectiveness across pipeline architectures and input configurations in broadcast-specific settings remains empirically undercharacterized. This paper presents a systematic evaluation of multimodal annotation pipelines applied to broadcast television news in the Italian setting. We construct a domain-specific benchmark of clips labeled across four semantic dimensions: visual environment classification, topic classification, sensitive content detection, and named entity recognition. Two different pipeline architectures are evaluated across nine frontier models, including Gemini 3.0 Pro, LLaMA 4 Maverick, Qwen-VL variants, and Gemma 3, under progressively enriched input strategies combining visual signals, automatic speech recognition, speaker diarization, and metadata. Experimental results demonstrate that gains from video input are strongly model-dependent: larger models effectively leverage temporal continuity, while smaller models show performance degradation under extended multimodal context, likely due to token overload. Beyond benchmarking, the selected pipeline is deployed on 14 full broadcast episodes, with minute-level annotations integrated with normalized audience measurement data provided by an Italian media company. This integration enables correlational analysis of topic-level audience sensitivity and generational engagement divergence, demonstrating the operational viability of the proposed framework for content-based audience analytics.

Conversation Tree Architecture: A Structured Framework for Context-Aware Multi-Branch LLM Conversations

Large language models (LLMs) are increasingly deployed for extended, multi-topic conversations, yet the flat, append-only structure of current conversation interfaces introduces a fundamental limitation: all context accumulates in a single unbounded window, causing topically distinct threads to bleed into one another and progressively degrade response quality. We term this failure mode logical context poisoning. In this paper, we introduce the Conversation Tree Architecture (CTA), a hierarchical framework that organizes LLM conversations as trees of discrete, context-isolated nodes. Each node maintains its own local context window; structured mechanisms govern how context flows between parent and child nodes, downstream on branch creation and upstream on branch deletion. We additionally introduce volatile nodes, transient branches whose local context must be selectively merged upward or permanently discarded before purging. We formalize the architecture's primitives, characterize the open design problems in context flow, relate our framework to prior work in LLM memory management, and describe a working prototype implementation. The CTA provides a principled foundation for structured conversational context management and extends naturally to multi-agent settings.