small language model

Automated surveillance of primary outcome reporting in clinical trials: Development and validation of a scalable screening pipeline.

Background: Detecting discrepancies between registered and published trial outcomes currently requires manual expert review, limiting surveillance to small samples. No scalable automated system exists for continuous monitoring of outcome reporting integrity across the clinical trial literature. We aimed to develop and...

Multimodal Transformer Integrating CT and Pathology for Postoperative Risk Stratification in Stage I Non-Small Cell Lung Cancer.

The DLRTransformer improved postoperative risk stratification compared with guideline-based risk stratification and identified patients with stage I NSCLC who benefited from adjuvant therapy.

Exploring the dimensions of autobiographical memory: a validation of the Autobiographical Recollection Test (ART) for the European-Portuguese language.

Effects of dialogic reading with Luganda primary 1 materials on children’s vocabulary and listening comprehension

Dialogic reading structures teacher–child interaction around shared texts to support oral language development, yet its effects when integrated with familiar language curriculum materials in Ugandan Primary 1 classrooms remain underexamined. This parallel cluster-randomized trial evaluated a 10-week dialogic reading ad...

SurgStreamer: proactive, request-driven monitoring of surgical video streams with structured state and a reusable reasoning layer

The results support separating perception from reasoning and point to perception under distribution shift as the main limitation.

Perception or Reasoning? Diagnosing and Tackling Reinforcement Learning Bottlenecks in VLMs

GRAD, a group-relative attention distillation objective that uses the best rollout in each group to directly guide the visual attention of weaker rollouts is proposed, which consistently improves performance across two model scales and seven benchmarks, and further complements existing RL methods.

Poisoning the Neighborhood: A Topology-Aware Topic-Level Attack on Large-Scale HNSW-based RAG Systems

Structural analyses show that the Hub-Satellite topology forms in HNSW, indicating that graph structure contributes to improved poisoned-document retrieval, and presents a topology-aware topic-level poisoning attack against large-scale HNSW-RAG that generalizes to unseen topic queries.

Visual Reasoning Failure Modes Protocol: construction and measurement of figure-based probes designed to separate careful human reading from vision-language model reading

When a vision-language model gets a question about a figure wrong, the score alone does not say why. The model may have missed the mark, taken a shortcut the figure offered, or answered from the question text without looking. Existing benchmarks control for one or two of these causes. The author has not found one that...

Externalized Intelligence and Model Substitution

The continuity of an artificial intelligence system’s memory and governance should not depend on retaining a particular large language model. This study investigates externalized intelligence as an architecture in which task-relevant memory, information selection, policy, and permissions are maintained outside model we...

Bridging the Human–AI Trust Gap: The Mechanisms Underlying Perceived Efficacy of AI-Powered Psychological Counseling Large Language Models

Artificial intelligence (AI) is transforming the delivery of mental health services, yet public trust in AI-based psychological counseling remains a significant challenge. Drawing on a “technology–context–trust” framework, this study recruited a total of 1000 consumers in China through Credamo and conducted five betwee...

Top Google Ads Agency in India: How to Choose the Right Partner

Finding the top Google Ads agency in India can help businesses improve their online visibility, generate quality leads, and increase sales. With more companies investing in paid advertising, choosing the right agency has become an important part of building a successful digital marketing strategy. However, the best age...

The Puzzle-Piece Fallacy: Why Materialism Collapses into Its Own Output

Modern physicalists are trapped inside a looping paradox of their own design. They consistently attempt to run reality in reverse, confusing the secondary output of a tool for the primary source of the toolmaker. The public exchanges compiled in this paper is a live audit of this systemic blind spot. When held to the g...