Ever heard of ChatGPT? What is "GPT" anyway? And why should businesses care? GPT (Generative Pre-trained Transformer) is transforming how we work by automating tasks and generating human-like text. Learn more about this game-changing AI in the thread below! #AI #B2B #Tech
LLMS
-
Unsloth TRL Rewrite: Improved Performance for AI Model Training
By
–
Thanks Daniel! I originally had the entire code in classic TRL and decided to rewrite it with Unsloth two days ago. I don't regret it, it's super performant and convenient. Great work! 🙂
-
MNP Leverages Databricks GenAI for Market Leadership
By
–
MNP, a leading Canadian professional services firm, wanted to strengthen its market leadership with GenAI. See how they leveraged Databricks to support the integration of LLMs and provide contextualized insights to their clients https://
dbricks.co/3LI6XTS -
GPT-4o Voice Mode Launches for Alpha Users This Week
By
–
En teoría esta semana empieza el acceso al modo voz de GPT-4o para algunos usuarios alpha, e imagino que paulatinamente irán sumándose cada vez más.
-
AI Model Collapse: Training on Synthetic Data Causes Gibberish
By
–
Beware of AI 'model collapse': How training on synthetic data pollutes the next generation Oxford scholars found that large language models fed a diet of 'cannibal' data, created by other LLMs, sink into complete gibberish. @AIatMeta @OpenAI @Nature @iliaishacked
-
Meta’s Llama 3.1 405B Fails Georgian Conjugation Task
By
–
I broke Meta's Llama 3.1 405B with one question (which GPT-4o mini gets right) The failure of the large language model may point to issues with the broad use of synthetic data. The table of conjugations is not actually the conjugation of the Georgian verb “ყოფნა.” It’s
-
Gemini 1.5 Pro tops AI safety harm classification leaderboards
By
–
We classified harms into High Harm and Low Harm. Congratulations to @GoogleDeepMind Gemini 1.5 Pro (post-IO) for being top of the list on both leaderboards!
-
Scale SEAL Leaderboard Evaluates AI Adversarial Robustness
By
–
1/ Scale is announcing our latest SEAL Leaderboard on Adversarial Robustness! Red team-generated prompts Focused on universal harm scenarios Transparent eval methods SEAL evals are private (not overfit), expert evals that refresh periodically http://
scale.com/leaderboard -
Multi-Step Tool Use in LLM Assistants for Complex Automation
By
–
The latest module from Cohere's LLM University, on Tool Use, is here! Explore how multi-step tool use enables complex task automations that require interaction with external systems. Build assistants that can plan, reason, and perform tool calls in a sequence.