AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

OSCAR: Online Soft Compression for RAG

Maxime Louis, Thibault Formal, Hervé Déjean, Stéphane Clinchant

ICLR 2026 · Large language models › Retrieval and knowledge · approach: RAG robustness and evaluation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
make compression query-dependent (ablation)OSCAR-llama (x16) · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.73 -> 0.77Improved
make compression query-dependent (ablation)OSCAR-llama (x128) · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.71 -> 0.75Improved
add compressor pretraining (ablation)OSCAR-llama · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.70 -> 0.77Improved
train the generator jointly instead of freezing it (ablation)OSCAR-llama · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.71 -> 0.77Improved
increase compression rate from x16 to x128OSCAR-llama · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.77 -> 0.75Hurt
use Llama-1B compressor instead of DeBERTa-v3 (ablation)OSCAR · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.73 -> 0.77Improved
use Llama-1B compressor instead of Modern-bert-large (ablation)OSCAR · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.76 -> 0.77Improved
switch retrieval from SPLADE+DeBERTa to BM25OSCAR-llama vs no compression · RAG QA (5 QA sets, LLM eval)average LLM-eval score: 0.58 vs 0.56No clear effect
use OSCAR-llama instead of no compressionMistral-24B · RAG QA, 10 docsaverage accuracy: 0.68 -> 0.69Improved
increase compressor depth from 8 to 10 layers (ablation)OSCAR-N-Layers · BEIRnDCG@10: 52.5 -> 54.3Improved
increase compressor depth from 10 to 16 layers (ablation)OSCAR-N-Layers · BEIRnDCG@10: 54.3 -> 55.3Improved
use e2e reranking by OSCAR itself instead of DeBERTa rerankerOSCAR-llama · RAG QA (6 QA sets, LLM eval)average LLM-eval score: 0.77 -> 0.77No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy