Semantic Thermodynamics – 79% LLM token reduction via narrative constraints

tauanvinicius1 pts0 comments

GitHub - tauansloboda/semantic-thermodynamics · GitHub

/" data-turbo-transient="true" />

Skip to content

Type / to search

Sign in<br>Sign upAppearance settings

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

tauansloboda

semantic-thermodynamics

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

main

BranchesTags

Go to file

CodeOpen more actions menu

Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit

History<br>1 Commit<br>1 Commit

DOCUMENTO FUNDACIONAL_ Termodinâmica Semântica v1.docx

DOCUMENTO FUNDACIONAL_ Termodinâmica Semântica v1.docx

README.md

README.md

Semantic_Thermodynamics_Whitepaper.pdf

Semantic_Thermodynamics_Whitepaper.pdf

experimento_um.py

experimento_um.py

experimento_um_resultados.csv

experimento_um_resultados.csv

experimento_zero.py

experimento_zero.py

experimento_zero_resultados.csv

experimento_zero_resultados.csv

View all files

Repository files navigation

Semantic Thermodynamics: LLM Entropy Minimization

This repository contains the foundational whitepaper, execution scripts, and raw empirical data for the Semantic Thermodynamics framework.

The Premise

Large Language Models operate as Bayesian inference engines. In environments with high semantic entropy, they expend excess compute mapping infinite phase spaces. By applying "Narrative Gravity"—a precise formula of persona, teleological vectors, and destructive pruning—we can force the model into a deterministic geodesic, drastically reducing cost and latency.

Empirical Results (Experiment Zero)

By restructuring a standard data-extraction prompt using the Formula v2.0 , we observed:

79.29% reduction in completion tokens.

60.73% reduction in system latency (from 3.4s to 1.3s).

Repository Contents

Semantic_Thermodynamics_Whitepaper.pdf: The complete theoretical framework and the Law of Entropic Proportionality ($\Lambda$).

experimento_zero.py: The async Python script used to benchmark the token and latency collapse.

experimento_um.py: The script mapping the "Structural Friction" and the optimal semantic gradient.

*.csv: Raw telemetry data from the OpenAI API runs.

How to Test the Physics

Clone this repo.

pip install openai

Export your API key: export OPENAI_API_KEY="sk-..."

Run python experimento_zero.py and watch the latency drop.

Author: Tauan Vinicius Guahyba Sloboda

About<br>No description, website, or topics provided.<br>Resources<br>Readme<br>Activity<br>Stars<br>0 stars<br>Watchers<br>0 watching<br>Forks<br>0 forks<br>Report repository

Releases

Packages

Contributors

Languages

You can’t perform that action at this time.

semantic thermodynamics commit data experimento_zero repository

Related Articles