ChatGPT is the way to go! Specially for bouncing ideas – gpt-5 is way more steerable than sonnet
LLMS
-
1B Parameter Model: 128K Context, Int4 Quantization, Llama 4 Distilled
By
–
chat is this real??? 128K context, int4 quantisation, 1B params, distilled from Llama 4
-

LLM-as-Judge Alignment Improves to 94% With Structured Evaluation
By
–
After refining our rubric design process, alignment between human reviewers and LLM-as-a-Judge improved from 37 % → 94 % — proof that structured evaluation enhances consistency.
-
Rubric-Driven Evaluation Improves LLM Alignment and Data Quality
By
–
Rubric-driven evaluation doesn’t just improve human and LLMaJ alignment. It accelerates delivery, reduces rework, and improves the thing we care about most: data quality.
-
Trusted Scale: Snorkel’s Framework for Scaling Trust in AI
By
–
This is Trusted Scale — Snorkel’s framework for designing, validating, and scaling trust in AI data pipelines. Read the full post by @pham_derek →
https://
snorkel.ai/blog/scaling-t
rust-rubrics-in-snorkels-quality-process/
… #AI #LLM #DataQuality #SnorkelAI #MachineLearning -

GPT-5 o3.1 iteration represents next leap in AI reasoning capabilities
By
–
Lead OpenAI researcher: "GPT-5, in some way, can be considered o3.1 – iteration of the same concept"
— Peter Gostev (@petergostev) 16 octobre 2025
"What I'm after right now is something next, what would be a significant jump to how we interact with models, that are more capable, think for even longer and interact with even… https://t.co/bZOtxQYXHYLead OpenAI researcher: "GPT-5, in some way, can be considered o3.1 – iteration of the same concept" "What I'm after right now is something next, what would be a significant jump to how we interact with models, that are more capable, think for even longer and interact with even
-
Poe Leaderboard Now Available for Desktop and Web
By
–
To view the current rankings, go to https://
poe.com/leaderboard (currently desktop and web only). (2/2) -

Poe Launches Daily Leaderboard for AI Models and Apps
By
–
New: Poe Leaderboard We launched a leaderboard that shows the most used AI models and apps across Poe's platform and API. Rankings are updated daily based on total token usage across the 200+ text, image, video, and audio bots on Poe. (1/2)
-
Building Skills Iteratively with Claude: Best Practices
By
–
If you want to make a Skill, we recommend building iteratively. Run Claude on real tasks, watch where it struggles, then capture what worked. Claude can help you write skills as well with our skill builder skill in claude dot ai.
-

Skills: Simple Folders for Specializing Claude Without Custom Agents
By
–
At their core, Skills are simple. They're just a folder with a http://
SKILL.md file. The file starts with a name and description, then contains instructions, code, and resources. This simplicity means anyone can now specialize Claude without building custom agents.