here's the OG diagram from google's "attention is all you need" (2017) โ the paper that started this whole AI wave
encoder left, decoder right, attention doing all the heavy lifting. every LLM today traces back to this exact architecture
wild that 8 boxes changed everything
made two LLMs debate over who came first โchicken or eggโ
gemma: โTHE EGG CAME FIRST. PERIOD.โ calls groqโs model a glorified autocomplete script
groqโs llama: calm, diplomatic, straight up therapist energy trying to de-escalate a group chat
itโs unreasonably funny
Hey @X algorithm
X
I'm looking to #connect with people interested in:
Agentic-AI
Frontend Backend
Full-stack
Al / ML
Data Science
JS/Frameworks
LeetCode & DSA
Freelancing
Building in public
If that's you, let's connect
you can fine-tune a model to be great at a task and simultaneously make it worse at explaining why it gave that answer
capability and interpretability donโt move together, sometimes they trade off
#llm#finetuning
n8n workflow status: alive, chaotic, occasionally profound
it tweeted a random thought about time being fake followed immediately by a fact about bees
no notes. 10/10. shipping this