AI Networking Cookbook: Practical recipes for AI-assisted network automation and development
Category
3,366 papers
My project, “A Contextual Memory Framework for Privacy-Preserving Local Cognitive Agents using SLM,” aims to develop an intelligent AI system that remembers previous conversations, user preferences, and important information across sessions while protecting user privacy. It uses a Small Language Model (SLM), structured...
Objectives Confirmation bias can distort information search and evidence interpretation in criminal investigations. We tested whether consulting a large language model (LLM) could reduce this bias in group decisions.Methods In total, 108 participants formed 36 three-person teams randomly assigned to LLM consultation, t...
A small point release that fixes install problems in 1.12.0 and one JSON Schema bug. Install fixes 1.12.0 could be installed alongside dependency versions it doesn't work with. Both problems show up as import errors. linkml-runtime must now be ≥ 1.12.0. linkml 1.12 uses linkml_runtime.utils.rdf_canonicalize, which firs...
Introduction Emotional intelligence (EI) is an individual-difference characteristic relevant to how health-professions students learn, cope, and engage with support, making its valid measurement a prerequisite for research on student development in this field. This study examined the psychometric performance and educat...
Reach audiences
Advertise in front of researchers, engineers, and readers.
Background-Students’ metacognitive awareness remains a challenge in biology learning, including the respiratory system. Objectives- This study aimed to develop AR-assisted ASICC worksheets and examine metacognitive awareness and skills. Methods-The study employed Educational Design Research using the Plomp model. Feasi...
Chain-of-thought prompting lets a language model write out its intermediate reasoning, and a wrong intermediate claim can carry through to a wrong final answer. This preprint measures that error propagation directly for Qwen2.5-1.5B, Qwen2.5-3B and Llama-3.2-3B Instruct, run locally in 4-bit. On ProofWriter, every grou...
Asked for the seahorse emoji, language models often supply a wrong one. Unicode has none, so the correct answer is a refusal. Using the Unicode registry as an exact oracle, we build 180 probe items (123 in an absent-emoji core) and 174 matched controls, scored under a five-category taxonomy. On Qwen2.5-Instruct at 0.5B...
Compressive strength is a key quality index of iron ore pellets, reflecting the combined effects of pellet integrity, material characteristics, thermal history, and operating conditions during grate–kiln pelletizing. Reliable early prediction of pellet compressive strength is therefore essential for burden-quality co...
Smaller language models often exhibit high confidence in erroneous predictions, undermining traditional ensemble methods. In heterogeneous ensembles, simple averaging can grant disproportionate influence to overconfident but incorrect outputs. This paper introduces a reliability-weighted extension of the Proportional C...
Pediatric speech and language assessment supports the early identification of communication difficulties and access to timely intervention. Large language models (LLMs) and related generative artificial intelligence technologies are increasingly being incorporated into systems that analyze children’s speech and languag...
Small language models offer a practical way to convert shopping queries into ex-ecutable constraints, but partial extraction accuracy can obscure whether the com-plete request is represented correctly. We study CommerceCore, a Qwen3-1.7B model adapted with QLoRA on 2,130 examples spanning entity segmentation, query con...
Purpose Large Language Models (LLMs) can be helpful for literature search and summarization but retracted articles can confuse them. It is therefore important to know how aware LLMs are of retractions. Design/methodology/approach Three open-weight (offline) LLMs were asked whether 161 high-profile retracted articles ha...
Generative AI
Assistant Professor Benjamin Lindquist studies how engineers deliberately injected randomness into early AI, giving rise to deep-learning models that sometimes hallucinate.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Flint-interactive gives AI agents a language for creating interactive visualizations. Click a chart, filter data, highlight related content, and explore beyond the view the author originally chose. The post Flint-interactive: A language for creating interactive data visualizations appeared first on Microsoft Research.
A Blog post by IBM Research on Hugging Face
A Blog post by IBM Research on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
