legal doc flow – receive legal doc
– upload to claude/o3
– ask it to create click to reveal plain english for sections
LLMS
-
Legal Document Workflow with Claude and Plain English
By
–
-
Upstage Selected for CB Insights AI 100 2025
By
–
🚀 Exciting News! @upstageai Selected for @CBinsights AI 100 in 2025! 🏆 This annual list celebrates the most promising private AI companies globally, and we're honored to be included. Our journey began in Korea, bringing advanced LLM technology to Thailand, and expanding to Japan and the U.S. This path has earned us global recognition, reinforcing our leadership in cutting-edge foundation models. Being part of the AI 100 highlights our commitment to innovation and our mission to shape the "Future of Work" through powerful AI solutions. A huge thank you to our amazing team, dedicated partners, and inspiring customers for making this achievement possible. We’re excited for the journey ahead and invite you all to continue supporting us as we grow on the global stage. Today is a great day! 🎉 #CBInsightsAI100 #Upstage #AIInnovation #AI2025 #SolarLLM
-
o4-Mini High Leads Coding Benchmarks Over Gemini 2.5 Flash
By
–
o4-Mini High reste le modèle leader en coding, loin devant ! Plus rapide, plus économique que les autres modèles OpenAI. Et info bonus intéressante : Gemini 2.5 Flash dépasse même légèrement le modèle Pro en coding ! Sans oublier que Sam a annoncé hier un doublement
-
AI Weekly Digest: Microsoft Agents, OpenAI Image API, Grok Updates
By
–
Ma dernière Newsletter quotidienne et gratuite est sortie. Voici les sujets que j'y aborde:
– Microsoft lance des agents d'IA sur le lieu de travail
– OpenAI lance une API de génération d'images
– Windsurf baisse ses prix face à la concurrence
– Grok ajoute la prise en -
Tree of Thought: AI Explores All Possible Reasoning Paths Simultaneously
By
–
Et si une IA ne suivait pas une pensée linéaire… mais toutes les pensées possibles à la fois ?
— VISION IA (@vision_ia) 24 avril 2025
L’arbre de pensée remplace la chaîne logique unique par une réflexion parallèle, exploratoire.
Résultat : plus de créativité, plus de robustesse.
C’est peut-être la clé pour des IA… pic.twitter.com/zIqt6N5cB7Et si une IA ne suivait pas une pensée linéaire… mais toutes les pensées possibles à la fois ? L’arbre de pensée remplace la chaîne logique unique par une réflexion parallèle, exploratoire. Résultat : plus de créativité, plus de robustesse. C’est peut-être la clé pour des IA
-
GPTs Boost Productivity Like a Personal Assistant
By
–
Great point! Using GPTs can really boost productivity. It's like having a personal assistant for all those boring tasks!
-
Tina: Open Source AI Project with Training Logs and Models
By
–
Notion Blog: https://
shangshangwang.notion.site/tina
Code: https://
github.com/shangshang-wan
g/Tina
…
Training Logs: https://
wandb.ai/upup-ashton-wa
ng-usc/Tina
…
Models: https://
huggingface.co/Tina-Yi -
Tiny Reasoning Models via LoRA Achieve Strong AIME Performance
By
–
Apparently, you can build tiny reasoning models via LoRA.
Tina is a family of 1.5B models trained using LoRA-based RL—cost-efficient and performant.
The best Tina hits >20% gain in reasoning performance and 43% Pass@1 on AIME24,
with just $9 in post-training + eval cost. -
Consensus Methods for Variance Reduction in LLMs
By
–
The interesting question is why and when would this work? Intuitively this looks like variance reduction in a setup where variance dominates bias because the LLM has very high capacity. So consensus makes sense. Thoughts @roydanroy @ArnaudDoucet1 ?? It has vibes of online
-
Prompt engineering is a skillset, not a job title
By
–
Prompt engineering isn’t a job title. It’s a skillset. It combines system design, language mastery, and product sense. And it’s becoming essential for anyone working with AI.