# CLaRa: Bridging Retrieval and Generation with Continuous Latent Reasoning

> RAG framework using continuous latent reasoning and embedding-based context compression

CLaRa is an AI research paper and framework by researchers from Apple and the University of Edinburgh that addresses RAG context-length and retrieval bottlenecks. It introduces continuous latent reasoning, compressing documents into dense memory vectors and training reranker and generator components end-to-end via a differentiable top-k estimator. It allows language models to maintain high retrieval and answer accuracy while feeding significantly fewer context tokens to the generator.

Website: https://arxiv.org/abs/2511.18659

## What it is for

- Compress long retrieved document contexts into compact latent vectors for LLM prompts
- Train retrieval rerankers and generators jointly via end-to-end language modeling loss
- Study advanced RAG architectures that move beyond discrete text-chunk retrieval

## Features

- Joint optimization of reranker and generator modules in a shared continuous space
- SCP data synthesis framework based on question answering and paraphrase supervision
- Differentiable top-k estimator enabling end-to-end gradient flow between retrieval and generation
- Text compression rates up to 16x without severe loss in downstream question-answering accuracy
- Shared continuous latent space aligning retrieval relevance directly with answer generation

## Pricing

Free

## Details

- Type: Article or course
- Category: [AI](https://xiry.app/topics/ai.md)
- Open source: no
- Platforms: Web
- For: AI researchers and machine learning engineers working on retrieval-augmented generation
- Alternative to: Standard RAG chunking pipelines, Dense passage retrieval (DPR)
- Tags: rag, research paper, embeddings, information retrieval, llm, context compression

## Why it was saved

- [@officeoptout, 15 December 2025](https://www.instagram.com/reel/DSSRfA8jNb-/): Solves traditional RAG semantic retrieval bottlenecks and reduces data size by compressing documents into dense memory tokens while training a query reasoner end-to-end to generate hypothetical answers before retrieval.

## Related

- [8 Ways to Improve RAG Applications](https://xiry.app/resources/8-ways-to-improve-rag-applications): Practical architectural strategies to fix common failure modes in RAG systems (Tip, Free)
- [Modern Information Retrieval Evaluation In The RAG Era](https://xiry.app/resources/modern-information-retrieval-evaluation-in-the-rag-era): Free video lecture on modern IR benchmarks and evaluation for RAG pipelines. (Article or course, Free)
- [AnythingLLM](https://xiry.app/resources/anythingllm): Private, on-device AI assistant to chat with documents and local models. (Desktop app, Open source)
- [Custom RAG Pipeline over Apple Notes](https://xiry.app/resources/custom-rag-pipeline-over-apple-notes): RAG architecture for chatting with local Apple Notes via DeepSeek (Tip, Free)
- [GPT4All](https://xiry.app/resources/gpt4all): Desktop app to run open-source language models locally and completely offline (Desktop app, Open source)
- [memary](https://xiry.app/resources/memary): Open-source persistent memory layer for autonomous AI agents using knowledge graphs (GitHub repo, Open source)

Updated 3 October 2026. Page: https://xiry.app/resources/clara-bridging-retrieval-and-generation-with-continuous-late

Make your own library: Xiry takes your saved reels, tweets, Reddit bookmarks and Watch Later videos, makes them searchable and pulls the useful stuff out, like this library. It's invite-only for now; join the waitlist at https://xiry.app/#waitlist.