Mitigating a language model's over-confidence with NLI predictions on Multi-NLI hypotheses with random word order using PAWS (paraphrase) and Winogrande (anaphora).

NLI Text Analytics Model to detect cross-linguistic text relationships

Data, models, analysis scripts, experiment materials and environments that I used for my MA thesis can be accessed in this repository.

Bachelor Thesis comparing two Relation-Inference Datasets

finetune the language model for task of natural language inference

Fine-tune transformers with NLI data

A vibe-coded news summarizer that combines Seq2Seq generation, NLI-based consistency checking, NER-driven entity tracking, and SBERT semantic embeddings to deliver context-aware summaries.

Fine-tuning ModernBERT for natural language inference (entailment / contradiction / neutral)

AI assistant system

Natural Language Inference, SNLI dataset

XNLIeu: a dataset for cross-lingual NLI in Basque

Re-implementation of BIMPM(Bilateral Multi-Perspective Matching for Natural Language Sentences)

This repository is an implementation of Paper "Robust Natural Language Inference Models with Example Forgetting", which studies importance of forgettable example in making robust NLI Models

Antibodies for LLMs hallucinations (grouping LLM as a judge, NLI, reward models)

This repository contains the code and models for the LegalLens-2024 shared task, which detects legal violations in unstructured text and associates them with affected individuals using transformer models like BERT, RoBERTa, and DeBERTa for Named Entity Recognition (L-NER) and Natural Language Inference (L-NLI).

Assignments and projects from the natural language processing course offered at the University of Tehran.

🌐 Multilingual NLI with BERT and Classical Models: A comparative study of traditional machine learning models and a fine-tuned BERT transformer for multilingual Natural Language Inference (NLI) using the "Contradictory, My Dear Watson" dataset. Demonstrates significant performance improvements with transformer-based approaches.

📃⚖️🙅 Pipeline to detect contradictions in policy documents using NLP and transformers.