Benchmarking Patent Embeddings: A Multi-Task Evaluation of 22 Models Across Retrieval, Classification, and Clustering
cs.IR, cs.AI
Submitted: 2026-05-22
Updated: 2026-11-05
Code: https://github.com/lightonai/pylate
Terminology
Sources
- PatenTEB: A Comprehensive Benchmark and Model Family for Patent Text Embedding
- M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation
- Large Language Models, Encoder Architectures and Hybrid Approaches for Patent Classification
- Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup
- PaECTER: Patent-level Representation Learning using Citation-informed Transformers
- Jina Embeddings 2: 8192-Token General-Purpose Text Embeddings for Long Documents
- KaLM-Embedding: Superior Training Data Brings A Stronger Embedding Model
- Towards General Text Embeddings with Multi-stage Contrastive Learning
- MTEB: Massive Text Embedding Benchmark
- Nomic Embed: Training a Reproducible Long Context Text Embedder
- ColBERTv2: Effective and Efficient Retrieval via Lightweight Late Interaction
- A comparative analysis of embedding models for patent similarity
- Text Embeddings by Weakly-Supervised Contrastive Pre-training
- IPBench: Benchmarking the Knowledge of Large Language Models in Intellectual Property
- PatentMind: A Multi-Aspect Reasoning Graph for Patent Similarity Evaluation
- Jasper and Stella: distillation of SOTA embedding models
- Patent Representation Learning via Self-supervision
Related papers
- UniRank: Benchmarking Ranking Models for Unified Sequential Modeling and Feature Interaction
- Revisiting Text Ranking in Deep Research
- GateSID: One Gate to Balance Semantic and Collaborative Signals across the Item Lifecycle
- SAM-D2Q: Aligning Multimodal Doc2Query with Search Demand and Conversion for E-commerce
- GINGER: Spicing Up Graph-Based Vector Search
- Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA