Entropic-Time Inference: Self-Organizing Large Language Model Decoding Beyond Attention
Fuente:
arXiv
Enregistré dans:
| Auteur principal: | |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910040003182592 |
|---|---|
| author | Kiruluta, Andrew |
| author_facet | Kiruluta, Andrew |
| contents | Modern large language model (LLM) inference engines optimize throughput and latency under fixed decoding rules, treating generation as a linear progression in token time. We propose a fundamentally different paradigm: entropic\-time inference, where decoding is governed by the flow of uncertainty rather than token index. We introduce a self\-organizing inference architecture that jointly couples scheduling, attention sparsification, and sampling temperature under a unified entropy control objective. Our method extends vLLM with entropy-aware scheduling, entropic pruning of paged attention blocks, and adaptive temperature control that stabilizes generation near a target entropy regime. This transforms inference into a resource\-intelligent thermodynamic process that allocates computation where uncertainty reduction is maximized. We present a concrete systems design, pseudocode, and integration plan, demonstrating how entropy can serve as a first\-class control signal for scalable LLM inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_03310 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Entropic-Time Inference: Self-Organizing Large Language Model Decoding Beyond Attention Kiruluta, Andrew Computation and Language Machine Learning Modern large language model (LLM) inference engines optimize throughput and latency under fixed decoding rules, treating generation as a linear progression in token time. We propose a fundamentally different paradigm: entropic\-time inference, where decoding is governed by the flow of uncertainty rather than token index. We introduce a self\-organizing inference architecture that jointly couples scheduling, attention sparsification, and sampling temperature under a unified entropy control objective. Our method extends vLLM with entropy-aware scheduling, entropic pruning of paged attention blocks, and adaptive temperature control that stabilizes generation near a target entropy regime. This transforms inference into a resource\-intelligent thermodynamic process that allocates computation where uncertainty reduction is maximized. We present a concrete systems design, pseudocode, and integration plan, demonstrating how entropy can serve as a first\-class control signal for scalable LLM inference. |
| title | Entropic-Time Inference: Self-Organizing Large Language Model Decoding Beyond Attention |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2603.03310 |