cody
538 posts
@cmcollier
Thinking and building ⌁ Always learning :: Making mistakes and moving forward
LLMs process text from left to right — each token can only look back at what came before it, never forward. This means that when you write a long prompt with context at the beginning and a question at the end, the model answers the question having "seen" the context, but the context tokens were generated without any awareness of what question was coming. This asymmetry is a basic structural property of how these models work. The paper asks what happens if you just send the prompt twice in a row, so that every part of the input gets a second pass where it can attend to every other part. The answer is that accuracy goes up across seven different benchmarks and seven different models (from the Gemini, ChatGPT, Claude, and DeepSeek series of LLMs), with no increase in the length of the model's output and no meaningful increase in response time — because processing the input is done in parallel by the hardware anyway. There are no new losses to compute, no finetuning, no clever prompt engineering beyond the repetition itself. The gap between this technique and doing nothing is sometimes small, sometimes large (one model went from 21% to 97% on a task involving finding a name in a list). If you are thinking about how to get better results from these models without paying for longer outputs or slower responses, that's a fairly concrete and low-effort finding. Read with AI tutor: chapterpal.com/s/1b15378b/pro… Get the PDF: arxiv.org/pdf/2512.14982
one thing llms don't change is that simplicity is the biggest unlock you can have when building software. and simplicity comes from a having a simple cognitive environment as well. how can you come up with simple solutions if you are orchestrating 100s of agents? I don't get it.
“when participants lost consciousness under anesthesia, … deep-brain structures became desynchronized.” Brain synchrony is a key candidate for the absorption of information into subjective experience.
I feel like I might finally understand what an engineer felt like during the era of the C compiler. The guy that wrote in Assembly language for quite a while, and then lived through the wave of C programming. Doing AI augmented programming, using natural language to drive code generation. It feels analogous to that historic crossing of an abstraction boundary.
