Pith Number

pith:TXQVCDGD

pith:2026:TXQVCDGDU5XNKGAJFHRL5YFFOO

not attested not anchored not stored refs resolved

Retrieval is Cheap, Show Me the Code: Executable Multi-Hop Reasoning for Retrieval-Augmented Generation

Ge Liu, Jash Rajesh Parekh, Jiajun Fan, Jiashuo Sun, Jiawei Han, Jimeng Shi, Peiran Li, Pengcheng Jiang, Qinglong Zheng, Saizhuo Wang, Shaowen Wang, Yixuan Xie, Zhiyi Shi

PyRAG reformulates multi-hop RAG as synthesis and execution of Python programs over retrieval tools.

arxiv:2605.12975 v1 · 2026-05-13 · cs.AI

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{TXQVCDGDU5XNKGAJFHRL5YFFOO}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

PyRAG reformulates multi-hop RAG as program synthesis and execution, enabling compiler-grounded self-repair and execution-driven adaptive retrieval without any additional training, and consistently outperforms strong baselines on five QA benchmarks with large gains on compositional multi-hop datasets.

C2weakest assumption

That code-specialized language models can reliably synthesize correct executable programs for multi-hop reasoning and that execution feedback alone suffices for self-repair without additional training or human oversight.

C3one line summary

PyRAG turns multi-hop reasoning into executable Python code over retrieval tools for explicit, verifiable step-by-step RAG.

References

51 extracted · 51 resolved · 13 Pith anchors

[1] Pathrag: Pruning graph-based retrieval augmented generation with relational paths 2026

[2] ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning 2025 · arXiv:2503.19470

[3] Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks 2022 · arXiv:2211.12588

[4] Binding language models in symbolic languages.arXiv preprint arXiv:2210.02875 2022

[5] From Local to Global: A Graph RAG Approach to Query-Focused Summarization 2024 · arXiv:2404.16130

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-18T03:09:08.763960Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

9de1510cc3a76ed5180929e2bee0a5738cae4cd717624087ade232a72852cc07

Aliases

arxiv: 2605.12975 · arxiv_version: 2605.12975v1 · doi: 10.48550/arxiv.2605.12975 · pith_short_12: TXQVCDGDU5XN · pith_short_16: TXQVCDGDU5XNKGAJ · pith_short_8: TXQVCDGD

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/TXQVCDGDU5XNKGAJFHRL5YFFOO \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 9de1510cc3a76ed5180929e2bee0a5738cae4cd717624087ade232a72852cc07

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "edd89fe4071ca30b1c3f0974ea426bf6988c50805ef42fad2f4913521ce95e34",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.AI",
    "submitted_at": "2026-05-13T04:14:13Z",
    "title_canon_sha256": "1d97c7f21586a0542518b277b13610e8eca1507d01b4dc4e6eb54a3a3817b6dd"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.12975",
    "kind": "arxiv",
    "version": 1
  }
}