Tootfinder

Modern applications require search capabilities that go beyond basic text matching. They must be fast, accurate, personalised and context-aware. At this year's Berlin Buzzwords, Saurabh Singh will demonstrate how OpenSearch’s latest AI/ML enhancements and engine improvements enable organisations to build intelligent, scalable search experiences that meet these evolving needs.
Learn more:

From Search to Insight: Leveraging OpenSearch for Scalable, AI-Driven Search Experiences
Observability, log analytics, GenAI systems, and RAG pipelines must query massive volumes of semantic embeddings to retrieve relevant content instantly. Today’s search systems often fall short in handling high-dimensional vector data and similarity search. OpenSearch 3.0 brings significant architectural improvements to address these challenges. Integrating Apache Lucene 10 and JVM 21, the platform delivers 20% faster queries than its 2.x predecessor and 10× the throughput of 1.x versions. Ne…

How to use the internet badly:
- I did a search on opening hours for the Scotrail ticket office in my town, got 6 different answers, 2 of them from Scotrail sources
- I tweeted to their twitter acct, telling them abt the variety of answers online, and could they please tell me the correct opening hours
- They tweeted back with what turned out to be incorrect information
- I thanked them and said they should contact the other travel sites and update their info /1

This https://arxiv.org/abs/2505.13626 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_…

Methods to Test the Source of the Extreme Gas Motions in WS 35
We present theoretical arguments toward the plausibility of a stellar wind}to explain the 16000\,km\,$s^{-1}$ line broadening in the optical spectra of WS 35, the central star in the Pa 30 nebula. The wind model is discussed in the context of super-Eddington flows. We argue that WS 35 potentially occupies a new regime of wind driving theory as the first metal-only wind. While this framework provides a promising avenue for explaining the high speed flow, questions remain about the source's true …

Good Morning #Canada
Yesterday, we purchased a new #HVAC system - a gas furnace and air-source #HeatPump. It's our 2nd heat pump as 16 years ago we installed geothermal in our previous home, thanks to a $10K energy rebate available at that time. Back then, I thought everyone would switch to this technology, but rebates disappeared, and upfront costs were a barrier. But new data shows heat pumps are almost even with traditional furnace installs.
#CanadaIsAwesome
https://440megatonnes.ca/insight/heat-pumps-continue-to-push-fossil-fuels-out-of-canadian-homes/

CLUES III: Do User Choices Impact The Results of SED Fitting? Tests of 'Off-The-Shelf' Stellar Population and Dust Extinction Models in the CLUES Sample
Andrew Mizener, Daniela Calzetti, Angela Adamo, Aida Wofford, Matthew J. Hayes, John Chisholm, Michele Fumagalli, Svea Hernandez, Matteo Maria Messa, Linda J. Smith, Arjan Bik, Kathryn Grasha, Mattia Sirressi

CLUES III: Do User Choices Impact The Results of SED Fitting? Tests of 'Off-The-Shelf' Stellar Population and Dust Extinction Models in the CLUES Sample
The simple stellar population models produced by stellar population and spectral synthesis (SPS) codes are used as spectral templates in a variety of astrophysical contexts. In this paper, we test the predictions of four commonly used stellar population synthesis codes (YGGDRASIL, BPASS, FSPS, and a modified form of GALAXEV which we call GALAXEVneb) by using them as spectral templates for photometric SED fitting with a sample of 18 young stellar clusters. All clusters have existing HST COS FUV …

Alas I've just ordered a WIFI enabled, AI washing machine; and the copyright doesn't even include CURL; I thought the 5 year warranty would enable be to ask @… for advice on getting my clothes cleaner.
Still, it seems to have everything else, NuttX and contiki low power OSs (neither of which I've come across), Uboot, wpa_supplicant, cJSON, FatFS

Evaluating Prompt Engineering Techniques for Accuracy and Confidence Elicitation in Medical LLMs
Nariman Naderi, Zahra Atf, Peter R Lewis, Aref Mahjoub far, Seyed Amir Ahmad Safavi-Naini, Ali Soroush
https://arxiv.org/abs/2506.00072

Evaluating Prompt Engineering Techniques for Accuracy and Confidence Elicitation in Medical LLMs
This paper investigates how prompt engineering techniques impact both accuracy and confidence elicitation in Large Language Models (LLMs) applied to medical contexts. Using a stratified dataset of Persian board exam questions across multiple specialties, we evaluated five LLMs - GPT-4o, o3-mini, Llama-3.3-70b, Llama-3.1-8b, and DeepSeek-v3 - across 156 configurations. These configurations varied in temperature settings (0.3, 0.7, 1.0), prompt styles (Chain-of-Thought, Few-Shot, Emotional, Exper…

Discover ColPali at Berlin Buzzwords 2025 with Sonam Pankaj. This session covers what ColPali is, how its "late-interaction" works, and how you can deploy its quantised version on your laptop.
Learn more: https://2025.berlinbuzzwords.de/sessio

Teaching Programming in the Age of Generative AI: Insights from Literature, Pedagogical Proposals, and Student Perspectives
Clemente Rubio-Manzano, Jazna Meza, Rodolfo Fernandez-Santibanez, Christian Vidal-Castro
https://arxiv.org/abs/2507.00108

Teaching Programming in the Age of Generative AI: Insights from Literature, Pedagogical Proposals, and Student Perspectives
Computer programming is undergoing a true transformation driven by powerful new tools for automatic source code generation based on large language models. This transformation is also manifesting in introductory programming courses at universities around the world, generating an in-depth debate about how programming content should be taught, learned, and assessed in the context of generative artificial intelligence. This article aims, on the one hand, to review the most relevant studies on thi…

Revisiting Chain-of-Thought Prompting: Zero-shot Can Be Stronger than Few-shot
Xiang Cheng, Chengyan Pan, Minjun Zhao, Deyang Li, Fangchao Liu, Xinyu Zhang, Xiao Zhang, Yong Liu
https://arxiv.org/abs/2506.14641

Revisiting Chain-of-Thought Prompting: Zero-shot Can Be Stronger than Few-shot
In-Context Learning (ICL) is an essential emergent ability of Large Language Models (LLMs), and recent studies introduce Chain-of-Thought (CoT) to exemplars of ICL to enhance the reasoning capability, especially in mathematics tasks. However, given the continuous advancement of model capabilities, it remains unclear whether CoT exemplars still benefit recent, stronger models in such tasks. Through systematic experiments, we find that for recent strong models such as the Qwen2.5 series, adding t…