PROMPT ENGINEERING

→ View original post on X — @rasbt

Flexibility in Reasoning LLMs with Preference Tuning

By

@rasbt

–

07 February 2025 22h27

Oh I see. Yeah maybe depends on how extensive the preference tuning was. But in general it should be possible to achieve the same "flexibility" there as with non-reasoning LLMs.

→ View original post on X — @mustafasuleyman

From Prompt Engineering to Personality Engineering: 2025 Trends

By

@mustafasuleyman

–

07 February 2025 19h12

2024: all about prompt engineering
2025: all about personality engineering

→ View original post on X — @rasbt

Wait as New Prompt Engineering Technique Replacing Step by Step

By

@rasbt

–

07 February 2025 15h01

"Wait." is the new "Think step by step"

Thinking Time Improves AI Model Accuracy

By

–

07 February 2025 10h35

7/ Thinking Time Improves Accuracy Research on deepseek-R1 confirms:
Longer response times = More accurate and insightful answers Encourage it with prompts like:
“Take your time and analyze this problem deeply before answering.”

Prompting strategies for DeepSeek-R1

By

–

07 February 2025 10h35

8/ Conclusion Deepseek-R1 is powerful, but it requires the right prompting strategies. Master these techniques and you’ll get far more intelligent responses. Try them and let me know how it goes!

Best Practices for Defining AI Output and Formatting

By

–

07 February 2025 10h35

6/ Define Output & Format Explicitly Being vague leads to poor results. Be very specific about the format. Example:
“Provide a detailed step-by-step breakdown in bullet points.” This forces structure and clarity.

Optimal Prompting Techniques for DeepSeek-R1

By

–

07 February 2025 10h35

5/ Zero-Shot Prompting Works Best Unlike other LLMs, deepseek-R1 typically performs better without examples. Instead of showing it a few answers, just use the techniques above for best results.