Semantic Thermodynamics – 79% LLM token reduction via narrative constraints
The Semantic Thermodynamics framework proposes using narrative constraints to reduce computational waste in large language models. Experiments demonstrate a 79.29% reduction in completion tokens and a 60.73% decrease in system latency when applying the Narrative Gravity formula. The repository provides the whitepaper, benchmarking scripts, and raw telemetry data for replication.
- ▪Large language models are described as Bayesian inference engines that expend excess compute in high semantic entropy environments.
- ▪Applying the Narrative Gravity formula, which combines persona, teleological vectors, and destructive pruning, forces the model into a deterministic geodesic, dramatically cutting token usage and latency.
- ▪Experiment Zero reported a 79.29% reduction in completion tokens and a latency drop from 3.4 seconds to 1.3 seconds.
- ▪The repository includes a whitepaper, Python benchmarking scripts, and CSV files containing raw OpenAI API telemetry data.
- ▪Instructions are provided to clone the repo, install dependencies, set an API key, and run the experiments to observe the performance gains.
Hacker News (AI / LLM) files mainly under ai. We currently carry 3,758 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | GitHub |
| Canonical URL | https://github.com/tauansloboda/semantic-thermodynamics |
| Publication time | Thu, 06 Aug 2026 00:00:41 +0000 |
| Retrieval time | 2026-08-06T00:05:43.584Z |
| Last seen | 2026-08-06T00:05:43.584Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | X3cG6Wn5581c · 1 stories |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
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).
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.