---
viewer: false
language:
- en
license: other
license_name: nlm-terms-and-conditions
license_link: https://www.nlm.nih.gov/databases/download/terms_and_conditions.html
task_categories:
- question-answering
tags:
- biomedical
- RAG
- BioASQ
- CLEF2026
- FAISS
- PISA
- BM25
- information-retrieval
pretty_name: Lean RAG Indexes for BioASQ Task 14b
size_categories:
- 10M<n<100M
---

# Lean RAG Indexes for BioASQ

**Paper:** [Retrieval-Bound Generation: Lean RAG Pipelines for Biomedical QA](https://clef-staging.pages.dev/paper6.pdf) — CLEF 2026 Working Notes, BioASQ Task 14b

**Code:** [github.com/lasigeBioTM/BioASQ14Taskb_2026](https://github.com/lasigeBioTM/BioASQ14Taskb_2026)

Prebuilt retrieval indexes for the **Lean RAG Pipelines for Biomedical Question Answering** project, developed as part of an MSc dissertation at LASIGE, University of Lisbon (in preparation).

These indexes support a hybrid (BM25 + dense retrieval) pipeline evaluated on [BioASQ Task 14b](http://bioasq.org/). Everything here was built to run on a single NVIDIA A30 GPU, so the configurations below favour resource efficiency and local deployability over peak retrieval performance.

## Repository Structure

```bash
faiss/
└── pubmed2026_ivfsq8_raw_matryoshka.index   # Dense vector index
pisa/
├── pubmed2026_soStopw/                      # BM25 index, stopword removal only
│   └── ...
└── pubmed2026_comDois/                      # BM25 index, stopword removal + Porter2 stemming
    └── ...
corpus/
└── pubmed2026.lmdb                          # Key-value store for document text
    └── ...
jsonl2026/
    └── ...                                  # PubMed 2026 Annual Baseline, JSONL format
```

## Corpus

All indexes are built over the **PubMed Annual Baseline 2026** (released 30 January 2026), comprising approximately 35 million documents. Only the title and abstract of each record are indexed.

<!-- TODO: fill in before publishing
- JSONL schema: which fields are present per line (e.g. pmid, title, abstract)?
- LMDB: what is used as the key, and what format is the stored value?
-->

## Indexes

### FAISS (Dense Retrieval)

| | |
|---|---|
| **File** | `faiss/pubmed2026_ivfsq8_raw_matryoshka.index` |
| **Index type** | IVF + Scalar Quantizer (SQ8) |
| **Clusters** | 2<sup>16</sup>, trained on a random sample of vectors from the corpus |
| **Similarity metric** | Inner product |
| **Embedding model** | [NeuML/pubmedbert-base-embeddings-matryoshka](https://huggingface.co/NeuML/pubmedbert-base-embeddings-matryoshka) |
| **Embedding dimension** | `768` |
| **Encoded text** | `title + abstract` |
| **Recommended `nprobe`** | `512` — tuned on the BioASQ training set for the best trade-off between Recall and per-query latency |

### PISA — Stopwords only (`pisa/pubmed2026_soStopw/`)

| | |
|---|---|
| **Retrieval model** | BM25 |
| **Preprocessing** | Stopword removal (PyTerrier / Terrier default stopword list) |
| **Stemming** | None |
| **Recommended BM25 parameters** | `k1=0.6`, `b=0.4` |

### PISA — Stopwords + Porter2 (`pisa/pubmed2026_comDois/`)

| | |
|---|---|
| **Retrieval model** | BM25 |
| **Preprocessing** | Stopword removal (PyTerrier / Terrier default stopword list) |
| **Stemming** | Porter2 |
| **Recommended BM25 parameters** | `k1=0.6`, `b=0.4` |

> **This is the configuration used for all reported results.** A grid search over `k1`, `b` and four preprocessing configurations (none, stemming only, stopwords only, both) was run on the BioASQ training set targeting Recall; stopword removal combined with Porter2 stemming came out on top. The `soStopw` index is included for ablation purposes.

**The BM25 parameters above are not PyTerrier's defaults** (`k1=1.2`, `b=0.75`). Using the defaults will not reproduce the reported results.

## Usage

### Download

```python
from huggingface_hub import snapshot_download

snapshot_download(
    repo_id="dantunes6/lean-rag-indexes",
    repo_type="dataset",
    local_dir="./lean-rag-indexes"
)
```

The full set of files is large. To fetch only what you need, use `allow_patterns`:

```python
snapshot_download(
    repo_id="dantunes6/lean-rag-indexes",
    repo_type="dataset",
    local_dir="./lean-rag-indexes",
    allow_patterns=["faiss/*"]
)
```

### Loading the FAISS index

```python
import faiss

index = faiss.read_index("faiss/pubmed2026_ivfsq8_raw_matryoshka.index")
index.nprobe = 512  # recommended
```

Queries must be encoded with the same model used to build the index
(`NeuML/pubmedbert-base-embeddings-matryoshka`) and normalised consistently with it,
since the index uses inner product as its similarity metric.

### Using the PISA index

The PISA index is used through [pyterrier_pisa](https://pyterrier.readthedocs.io/en/stable/ext/pyterrier_pisa/). Point your retriever at the index directory after downloading, and set the BM25 parameters explicitly:

```python
from pyterrier_pisa import PisaIndex

index = PisaIndex("pisa/pubmed2026_comDois")
bm25 = index.bm25(k1=0.6, b=0.4, num_results=1000)
```

## Corpus Notice

The corpus used to build these indexes consists of PubMed abstracts from the **PubMed Annual Baseline 2026**, distributed by the [National Library of Medicine (NLM)](https://pubmed.ncbi.nlm.nih.gov/download/).

NLM does not endorse or claim any responsibility for derived works, and the data is not covered by copyright in the United States — but records may contain material subject to third-party rights, and NLM's [Terms and Conditions](https://www.nlm.nih.gov/databases/download/terms_and_conditions.html) apply. Please make sure you comply with them before redistributing this data.

## Citation

If you use these indexes, please cite the working notes:

```bibtex
@inproceedings{antunes2026retrievalbound,
  author    = {Antunes, Diogo M. C. and Lopes, Paulo R. C. and Couto, Francisco M.},
  title     = {Retrieval-Bound Generation: Lean {RAG} Pipelines for Biomedical {QA}},
  booktitle = {Working Notes of the Conference and Labs of the Evaluation Forum ({CLEF} 2026)},
  series    = {CEUR Workshop Proceedings},
  publisher = {CEUR-WS.org},
  address   = {Jena, Germany},
  pages     = {123--145},
  year      = {2026}
}
```

*The permanent CEUR-WS volume number and URL will be added once the proceedings are published (September 2026).*

## Credits

Developed at [LASIGE](https://lasige.pt/), University of Lisbon, by Diogo Antunes.
Supervised by Francisco M. Couto.

## PS

I am continuously looking to improve the quality of this dataset. If you spot any inconsistencies, errors, or anomalies in the data, please feel free to report them by opening a new discussion in the Community tab. Your feedback is highly appreciated!