@wzhao_nlp is the expert on this and she said she thinks they have a classifier trained for refusal, probably just with some supervised data it would still work well even if it were pretty small (<1B params)
PROMPT ENGINEERING
-
Promptly: Learning AI Prompting Skills Through Gamified Feedback
By
–
Third prize went to Promptly: a tool for learning effective AI prompting through games that evaluate prompts. Users play pong and minesweeper by prompting Claude, and then receive feedback to improve their prompting skills.
-
ChatGPT Web Browsing Reluctance Issues Remain Unresolved
By
–
ChatGPT web browsing is still extremely reluctant to actually browse the web…
-
Prompt Engineering Techniques for Consistent LLM Results
By
–
Use prompt engineering to tailor your LLMS to your needs. Use reliable system prompts, few-shot prompting, or chain-of-thought prompting for desirable, consistent LLM results.
-
Mastering LLM Output Control and Model Parameter Tuning
By
–
That's just the tip of the iceberg! To dig deeper into controlling your LLM outputs and optimally tuning your model's parameters, expand your knowledge with this newsletter iteration: https://
louisbouchard.substack.com/p/our-ai-tutor
-is-live
… Subscribe to the newsletter for more weekly AI insights like these! -
Stop Sequences for Controlling LLM Output Length and Structure
By
–
Stop sequences aid in managing the length and structure of model-generated text, useful when prompts require reasoning and referencing. This is key in the React framework where LLMs tackle simpler tasks sequentially to answer complex prompts.
-
Model Behavior Optimization Through Examples and Parameter Tuning
By
–
Identify your model's desired behavior. Create user-model interaction examples to assess its performance and identify improvement areas. Enhancements can be achieved through adjusting decoding parameters.
-
Temperature Parameter: Controlling Randomness and Creativity in AI Models
By
–
Temperature in a model dictates randomness: below 1 yields deterministic predictions, above 1 increases variability. Lower temperatures aid summarization, while higher ones enhance creativity.
-
Controlling LLM Outputs: Techniques to Improve Performance
By
–
Using Large Language Models like GPT-4 and Llama 2 may be challenging, but with proper techniques, their outputs can be controlled effectively. A variety of measures exist to improve these models' performance.
-
Master Prompt Engineering Skills in GenAI Webinar November 28
By
–
Uncover #GenAI magic on Nov 28 in our webinar. Dive into prompt crafting, streamline LLM interactions, and more. Don't miss out on leveling up your Prompt Engineering skills. Register now! https://
domino.buzz/3MlpW78 #Webinar #AI