Livnium v3: Multi-Scale Attractor Dynamics, Token-Level Alignment, and Divergence-Based Reliability for Natural Language Inference

Fuente: Zenodo
Guardado en:
Detalles Bibliográficos
Autor principal: Patil, Chetan
Formato: Recurso digital
Lenguaje:inglés
Publicado: Zenodo 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866901168222896128
author Patil, Chetan
author_facet Patil, Chetan
contents <p>We extend the Livnium attractor-based NLI system with three contributions: (1) a cross-encoder upgrade improving SNLI dev accuracy from 82.2% to 84.5% via joint [CLS] premise [SEP] hypothesis [SEP] encoding; (2) token-level alignment extraction from the last-layer BERT cross-attention block, rendering the model's internal structural comparison visible as a force map between premise and hypothesis tokens; and (3) an alignment divergence metric that serves as a zero-cost intrinsic reliability signal, with empirically validated thresholds distinguishing stable from unreliable predictions. We further demonstrate that the same constraint-injection mechanism reproduces the Bayesian belief update in the Monty Hall problem, connecting NLI inference to classical decision theory through a unified energy-reshaping framework.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_19433529
institution Zenodo
language eng
publishDate 2026
publisher Zenodo
record_format zenodo
spellingShingle Livnium v3: Multi-Scale Attractor Dynamics, Token-Level Alignment, and Divergence-Based Reliability for Natural Language Inference
Patil, Chetan
natural language inference
NLI
attractor dynamics
BERT
cross-encoder
attention alignment
divergence
reliability
Monty Hall
energy-based models
SNLI
<p>We extend the Livnium attractor-based NLI system with three contributions: (1) a cross-encoder upgrade improving SNLI dev accuracy from 82.2% to 84.5% via joint [CLS] premise [SEP] hypothesis [SEP] encoding; (2) token-level alignment extraction from the last-layer BERT cross-attention block, rendering the model's internal structural comparison visible as a force map between premise and hypothesis tokens; and (3) an alignment divergence metric that serves as a zero-cost intrinsic reliability signal, with empirically validated thresholds distinguishing stable from unreliable predictions. We further demonstrate that the same constraint-injection mechanism reproduces the Bayesian belief update in the Monty Hall problem, connecting NLI inference to classical decision theory through a unified energy-reshaping framework.</p>
title Livnium v3: Multi-Scale Attractor Dynamics, Token-Level Alignment, and Divergence-Based Reliability for Natural Language Inference
topic natural language inference
NLI
attractor dynamics
BERT
cross-encoder
attention alignment
divergence
reliability
Monty Hall
energy-based models
SNLI
url https://doi.org/10.5281/zenodo.19433529