eric11eca/NeuralLog
A neural-symbolic joint reasoning approach for Natural Language Inference (NLI). Modeling NLI as inference path planning through a search engine. Sequence chunking and neural paraphrase detection for syntactic variation. SOTA result on SICK and MED.
This project helps evaluate if one natural language statement logically follows from another. You input two sentences, and it determines the logical relationship between them: entailment, contradiction, or neutral. This is useful for researchers and developers working on advanced text understanding and AI reasoning systems.
No commits in the last 6 months.
Use this if you need to build or benchmark systems that require highly accurate logical deduction from text, combining both statistical patterns and explicit logical rules.
Not ideal if you are looking for a simple, off-the-shelf tool for general text summarization or sentiment analysis without the need for deep logical inference.
Stars
16
Forks
3
Language
Jupyter Notebook
License
MIT
Category
Last pushed
Jun 09, 2021
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/eric11eca/NeuralLog"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
coetaur0/ESIM
Implementation of the ESIM model for natural language inference with PyTorch
erickrf/multiffn-nli
Implementation of the multi feed-forward network architecture by Parikh et al. (2016) for...
vanzytay/EMNLP2018_NLI
Repository for NLI models (EMNLP 2018)
hsinyuan-huang/FusionNet-NLI
An example for applying FusionNet to Natural Language Inference
sdnr1/EBIM-NLI
Enhanced BiLSTM Inference Model for Natural Language Inference