EPISODE · Mar 13, 2026 · 24 MIN
LLM2Vec-Gen: Generative Embeddings from Large Language Models
from Daily Paper Cast · host Jingwen Liang, Gengyu Wang
🤗 Upvotes: 23 | cs.CL Authors: Parishad BehnamGhader, Vaibhav Adlakha, Fabian David Schmidt, Nicolas Chapados, Marius Mosbach, Siva Reddy Title: LLM2Vec-Gen: Generative Embeddings from Large Language Models Arxiv: http://arxiv.org/abs/2603.10913v1 Abstract: LLM-based text embedders typically encode the semantic content of their input. However, embedding tasks require mapping diverse inputs to similar outputs. Typically, this input-output is addressed by training embedding models with paired data using contrastive learning. In this work, we propose a novel self-supervised approach, LLM2Vec-Gen, which adopts a different paradigm: rather than encoding the input, we learn to represent the model's potential response. Specifically, we add trainable special tokens to the LLM's vocabulary, append them to input, and optimize them to represent the LLM's response in a fixed-length sequence. Training is guided by the LLM's own completion for the query, along with an unsupervised embedding teacher that provides distillation targets. This formulation helps to bridge the input-output gap and transfers LLM capabilities such as safety alignment and reasoning to embedding tasks. Crucially, the LLM backbone remains frozen and training requires only unlabeled queries. LLM2Vec-Gen achieves state-of-the-art self-supervised performance on the Massive Text Embedding Benchmark (MTEB), improving by 9.3% over the best unsupervised embedding teacher. We also observe up to 43.2% reduction in harmful content retrieval and 29.3% improvement in reasoning capabilities for embedding tasks. Finally, the learned embeddings are interpretable and can be decoded into text to reveal their semantic content.
What this episode covers
🤗 Upvotes: 23 | cs.CL Authors: Parishad BehnamGhader, Vaibhav Adlakha, Fabian David Schmidt, Nicolas Chapados, Marius Mosbach, Siva Reddy Title: LLM2Vec-Gen: Generative Embeddings from Large Language Models Arxiv: http://arxiv.org/abs/2603.10913v1 Abstract: LLM-based text embedders typically encode the semantic content of their input. However, embedding tasks require mapping diverse inputs to similar outputs. Typically, this input-output is addressed by training embedding models with paired data using contrastive learning. In this work, we propose a novel self-supervised approach, LLM2Vec-Gen, which adopts a different paradigm: rather than encoding the input, we learn to represent the model's potential response. Specifically, we add trainable special tokens to the LLM's vocabulary, append them to input, and optimize them to represent the LLM's response in a fixed-length sequence. Training is guided by the LLM's own completion for the query, along with an unsupervised embedding teacher that provides distillation targets. This formulation helps to bridge the input-output gap and transfers LLM capabilities such as safety alignment and reasoning to embedding tasks. Crucially, the LLM backbone remains frozen and training requires only unlabeled queries. LLM2Vec-Gen achieves state-of-the-art self-supervised performance on the Massive Text Embedding Benchmark (MTEB), improving by 9.3% over the best unsupervised embedding teacher. We also observe up to 43.2% reduction in harmful content retrieval and 29.3% improvement in reasoning capabilities for embedding tasks. Finally, the learned embeddings are interpretable and can be decoded into text to reveal their semantic content.
NOW PLAYING
LLM2Vec-Gen: Generative Embeddings from Large Language Models
No transcript for this episode yet
Similar Episodes
No similar episodes found.