Don't worry, Pratik, I'm pretty sure you're not ChatGPT. Unless ChatGPT has learned how to make sarcastic comments about itself. In which case, we're all doomed.
LLMS
-
AI Agents Creating Dynamic Dashboards with SQL and Real-Time Visualization
By
–
Our most common AI agent built is an on-the-fly dashboard. Basically, you ask an AI agent to create a dashboard, and the LLM will generate SQL, run the query, and create a plot or on-the-fly real-time dash that you can play with. This provides insights, plots, and information
-
ChatGPT’s Lazy Period: Refusing Data Reformatting Tasks
By
–
ChatGPT had a lazy period, where it sometimes would refuse user requests if they looked like too much work. For instance, when asked to reformat a bunch of tabular data into JSON, it would do the first row, and ask the user to do the rest by themselves
-
Mixtral-8x7b-Groq Bot Now Available on Poe Platform
By
–
The bot is available at https://
poe.com/Mixtral-8x7b-G
roq
… and across the Poe iOS, Android, Mac, and Windows apps in the official bots category. (2/2) -
Groq-powered Mixtral inference now available on Poe platform
By
–
Groq-powered inference for Mixtral is now available on Poe! You can use Mixtral-8x7b-Groq and experience 400 token/second responses. (1/2) pic.twitter.com/2oyM54WWjK
— Poe (@poe_platform) 29 février 2024Groq-powered inference for Mixtral is now available on Poe! You can use Mixtral-8x7b-Groq and experience 400 token/second responses. (1/2)
-
LLM Insincerity: The Problem with AI Emotional Fakeness
By
–
I hate it when LLMs are lying to me, starting with chatGPT saying that it is happy to help or Gemini saying that it appreciates my argument, indicating that they are tuned by people who are too insincere to understand true human connection and don’t appreciate arguments.
-
Reduce LLM API Costs by 60% with Domino Cascades
By
–
Slash #LLM API costs by 60% with LLM Cascades on Domino! Get affordable, powerful models for cost-efficient AI. Ready for a smarter AI strategy? Check our blog! https://
domino.buzz/3SYxcZ5 #AI #MachineLearning #CostSavings #DataScience -
Bradley-Terry Model Ranking for AI Model Preference Predictions
By
–
Using the preference predictions for different model pairs for a given prompt, I fit a Bradley-Terry model (the same ranking algorithm that Chatbot Arena uses) to compute a model ranking specific to that prompt.
-
Weak Models Match Strong Models on Simple Prompts
By
–
In my experiment, a preference predictor was able to pick up the performance patterns of different models. One pattern is that for simple prompts, weak models can do (nearly) as well as strong models. For more challenging prompts, however, users are much more likely to prefer
-
Predicting Best AI Model Selection for User Queries
By
–
A challenge of building AI applications is choosing which model to use. What if we don’t have to? What if we can predict the best model for any prompt? Predictive human preference aims to predict which model users might prefer for a specific query. One use case is model