bi-encoder

Sentence-Transformers Information Retrieval example on Chinese

Code and created datasets for our ACL 2022 paper: "Contextual Fine-to-Coarse Distillation for Coarse-grained Response Selection in Open-Domain Conversations"

LoRA fine-tuning of bi-encoder retrievers with hard negatives and cross-encoder distillation, evaluated on NFCorpus.

ViIR: The Unified Framework for Fine-tuning Vietnamese Information Retrieval Models with Various Tuning Statergies.

High-accuracy job classification system using Sentence Transformers. Maps job titles & descriptions to 1,016 O*NET-SOC categories. 100% Top-1 accuracy on real job postings. Fast CPU inference (<100ms). 126K+ training samples from 8 O*NET data sources.

Explainable AI resume-screening engine — ranks candidates against a job description and shows why each scored as it did. FastAPI · Sentence-BERT · HyRe · Groq.

Official codebase for the ACL 2026 MeLLM Workshop paper "The Multilingual Curse at the Retrieval Layer: Evidence from Amharic"

Full-corpus SemCor relabeling with GPT-5.5 and Gemma, controlled BEM/ESCHER/ConSeC retraining, and the Glite LENS bi-encoder - reproducibility record for the lexEn/SenseBench WSD paper

Powered by a catalog of 190+ products, this engine delivers high-precision results using semantic embeddings and vector similarity principles. By mapping product data into high-dimensional space and calculating the cosine similarity between search queries and items, it identifies matches based on intent and meaning rather than just keywords.

Comparative study of parameter-efficient fine-tuning (PEFT) strategies for biomedical NER on top of GLiNER — including soft prompt tuning, embedding injection, and a custom in-place embedding extension that matches full fine-tuning performance at 13% of trainable parameters.

Proof of concept for large language model summarization of medical journal articles for different reading levels

Baseline models for searching for movie plots from Wikipedia articles. Techniques include BM25 (lexical search), bi/cross-encoding (semantic search), and retrieval-augmented generation (RAG) using Mistal 7B through Fireworks.ai.

Neural search engine with bi-encoder + cross-encoder re-ranking, BM25 hybrid search, query expansion, typo correction and multi-language support — processes 10M queries/day

Évaluation de la pertinence (question ↔ article juridique) en français. Pipeline complet (prépa → modèles → soumission) avec CamemBERT en bi-encodeur calibré (MSE/Spearman), + variantes cross-encoder.

A production-grade, two-stage Financial RAG architecture analyzing SEC 10-K reports. Features an evaluation sandbox (RAGAS) and an asynchronous, containerized FastAPI backend powered by Gemini Flash and ChromaDB.

InsureLLM RAG Challenge — Two-stage retrieval pipeline (Bi-Encoder + Cross-Encoder) with context compression

Tiny semantic search engine with bi-encoder retrieval and cross-encoder reranking — built to understand how production search works under the hood.

A self-hosted AI-powered RSS aggregator that uses bi-encoders and LLMs to semantically filter daily articles based on a custom prompt.

Recommendation systems overview and an MLOps TFX-pipeline implementation

Exploring fast & accurate zero-shot text classification