Fibonacci Language Model Experiment: Analyzing Word Patterns at Fibonacci Intervals
By
gabriel666smith
Crackling crust, pillowy middle. The kind of bagel that earns a second cup of coffee.
Summary
A developer describes an experimental project where they used Claude (an AI assistant) to write a script analyzing 4MB of prose text to find words that appear at Fibonacci intervals (both forwards and backwards). The output was a 100MB model file. The author expresses uncertainty about whether the results are meaningful and seeks help understanding if the approach has any validity as a language model.
Key quotes
· 5 pulledFML LM: or, FML, this sort of looks like a Fibonacci Language Model.
Can anyone help me work out if it is meaningful?
I took 4mb of text data from a selection of prose that I thought was interesting. I'm not sure if the results are or not.
I got Claude to write a script that looked for the words within that 4mb of text data which appeared, most often, at Fibonacci intervals to each individual word, both forwards and backwards in the text.
The output of the script was a 100mb pkg model.
You might also wanna read
NVIDIA Announces "Hack for Impact" London Event for Autonomous AI Agent Development
NVIDIA is hosting a "Hack for Impact" event in London, challenging participants to build autonomous agentic applications using open-source m
luma.com·38m ago
Four practical steps to control Azure Foundry token costs for agentic AI workloads
This article provides practical guidance on controlling token costs in Microsoft Azure Foundry, particularly for agentic AI workloads where
purplefrogsystems.com·1h ago
MerLean-Prover: A Recursive Agent Harness for Lean 4 Theorem Proving Outperforms Baselines
MerLean-Prover is an end-to-end Lean4 theorem prover that replaces 'sorry' declarations with kernel-checkable proofs using three agent types
arxiv.org·1h ago
Why small pull request policies can backfire on software quality
The article critiques a common software engineering policy that limits pull requests (PRs) to small sizes (e.g., 500 lines, few files). Whil
apenwarr.ca·3h ago
How Anthropic contains Claude's expanding access across its products
Anthropic describes how it has evolved its approach to granting Claude, its AI assistant, increasingly broad access to internal systems over
anthropic.com·5h ago
Testing Cursor's Jira integration: How ticket quality affects AI agent performance
Cursor launched a Jira integration that lets developers assign tickets directly to an AI agent, eliminating context switching. The author te
bit.ly·5h ago