Obvio la parte más interesante del leak, el listado de verbos inventados que se muestran cuando Claude piensa
PROMPT ENGINEERING
-
Model Training with SFT Using Claude and GPT-4o-mini
By
–
When I published this post I hadn't yet come across Trip's own writeup of the project, which provides way more detail about how he trained the model including SFT (supervised fine tuning) against synthetic chat examples created using Claude Haiku and GPT-4o-mini
-
First Conference with SKILL.md Submissions
By
–
The first conference where your submission isn't a paper.
— God of Prompt (@godofprompt) 31 mars 2026
It's a SKILL.md file Claude can execute, review, and reproduce end-to-end.
Stanford and Princeton just made reproducibility the submission format.
$50,000 prize pool. 364 winners.
Deadline April 5 👇… https://t.co/4ql4fVGe4QThe first conference where your submission isn't a paper. It's a SKILL.md file Claude can execute, review, and reproduce end-to-end. Stanford and Princeton just made reproducibility the submission format. $50,000 prize pool. 364 winners. Deadline April 5
-
Prompt Quality is 80% of AI Results
By
–
On sous-estime tellement la qualité du prompt… alors que c’est 80% du résultat.
-
Evolutionary Prompt Optimization vs Direct Model Iteration
By
–
Evolutionary prompt optimization is super interesting. How does it compare to just iterating with the model directly? 🙂
-
Multi-model routing for deep research systems
By
–
Multi-model for deep research makes sense. Curious how it picks which model for what, or if users control the routing.
-
10 Token-Saving Hacks to Maximize Claude Usage
By
–
Here's the ultimate cheat sheet to never hitting Claude's limits again Save tokens with these 10 hacks: #1 Rewrite, don't follow up
#2 Cut history with new chats
#3 Merge your prompts
#4 Use Projects to cache docs
#5 Set default User Memory
#6 Deactivate Search
#7 Route easy -
Meta-Harness: Optimizing LLM Model Harnesses End-to-End
By
–
"Meta-Harness: End-to-End Optimization of Model Harnesses" A big portion of agentic LLM performance relies on human-designed harness around the model, not just the weights. On top of that, how well a harness is designed could impact a model's performance heavily. So this paper
-
New Practical Guide to Reinforcement Learning from Human Feedback Released
By
–
New release from @PacktDataML available at http://
amzn.to/3PMn1ZL A Practical Guide to Reinforcement Learning from Human Feedback (RLHF). Amazon Summary: RLHF is a powerful approach to AI alignment and human-centered machine learning. By combining reinforcement learning -
Auto Compaction in Codex CLI: Context Window Management Feedback
By
–
You can use /compact in Codex CLI, but people really like the auto compaction in Codex and tend to forget about the context window. Would love your feedback!