https://peertube.watch/videos/watch/ca9eb527-1c73-4d37-804f-10239feab447 Peace Call. Mr. Peace is an AI-made character; voice is AI-generated; script written with AI assistance. No affiliate links in this video. How this site is run: https://getpeacecall.com/verify/?utm_source=peertube&utm_medium=video&utm_campaign=underdog-relex In a preprint by Christou and Tsoumakas, a fine-tuned Qwen2.5-0.5B (2 in-context examples) averaged 0.828 F1 over 7 general-domain relation-extraction datasets, vs 0.693 for zero-shot GPT-5.4 and 0.662 for zero-shot Claude Sonnet 4.6 (Table 4). A 3B model, Llama-3.2-3B, topped the table at 0.844. In this study, the tuned two-shot model outscored zero-shot GPT and Claude. That doesn't establish overall model superiority; the authors say this comparison does not imply small models are intrinsically stronger. Same base model as our earlier 0.5B vs 72B video - a different study. Source: Christou & Tsoumakas (Aristotle University of Thessaloniki), 'Sub-Billion, Super-Frontier: Small Language Models Rival Zero-Shot Frontier LLMs on General and Literary Relation Extraction', arXiv:2606.22606v1 - PREPRINT (Jun 21, 2026), not yet peer reviewed. https://arxiv.org/abs/2606.22606 . Positive-class micro-F1, dataset-macro average over 7 benchmarks. Tuned Qwen 2-shot vs zero-shot GPT/Claude. Names are trademarks of their respective owners; Peace Call is not affiliated with or endorsed by them. Charts drawn by code from the source's numbers. Paper: https://arxiv.org/abs/2606.22606 (version v1, posted Jun 21, 2026). AI Disclosure: Mr. Peace is a fictional AI character using a synthetic voice. This video is AI-assisted and does not depict a real person speaking. Research Disclosure: Based on a research preprint that has not been peer-reviewed. Results apply to the reported benchmarks and experimental conditions, not overall model superiority. Trademark Disclosure: Names are trademarks of their respective owners. Peace Call is not affiliated with or endorsed by them. Thu, 05 Nov 2026 03:49:56 GMT https://validator.w3.org/feed/docs/rss2.html PeerTube - https://peertube.watch https://peertube.watch/lazy-static/avatars/3b284627-b75f-409e-ac2e-10f4c2946f43.png https://peertube.watch/videos/watch/ca9eb527-1c73-4d37-804f-10239feab447 All rights reserved, unless otherwise specified in the terms specified at https://peertube.watch/about and potential licenses granted by each content's rightholder.