Pith Number

pith:WTB52ZRB

pith:2026:WTB52ZRBMY3ZABR43EZHIM2ZMO

not attested not anchored not stored refs resolved

Generative Floor Plan Design with LLMs via Reinforcement Learning with Verifiable Rewards

Aditya Sharma, Aristides Milios, Christopher Beckham, Christopher Pal, Florian Golemo, Ge Ya Luo, Luis Lara, Zhi Hao Luo

Fine-tuning an LLM then applying reinforcement learning with verifiable rewards produces floor plans that meet both room connectivity and exact numerical constraints on dimensions and areas.

arxiv:2605.14117 v1 · 2026-05-13 · cs.CL · 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{WTB52ZRBMY3ZABR43EZHIM2ZMO}

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

Our model generates floor plans that satisfy user-defined connectivity and numerical constraints and outperforms existing methods on Realism, Compatibility, and Diversity metrics. Across all tasks, our approach achieves at least a 94% relative reduction in Compatibility compared with existing methods.

C2weakest assumption

The verifiable rewards and constraint adherence metrics fully capture functional validity of floor plans without overlooking buildability, safety, or aesthetic factors that matter in real professional use.

C3one line summary

Fine-tuned LLMs trained with reinforcement learning using verifiable rewards produce floor plans that satisfy connectivity and numerical constraints, outperforming prior methods with at least 94% relative improvement in compatibility.

References

16 extracted · 16 resolved · 5 Pith anchors

[1] ArXiv:2506.14702v1

[2] Architext: Language-driven generative architecture design.arXiv preprint arXiv:2303.07519. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al- Dahle, Aiesha

[3] The Llama 3 Herd of Models · arXiv:2407.21783

[4] LoRA: Low-Rank Adaptation of Large Language Models · arXiv:2106.09685

[5] Efficient Memory Management for Large Language Model Serving with PagedAttention · arXiv:2309.06180

Receipt and verification

First computed	2026-05-17T23:39:11.937158Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

b4c3dd6621663790063cd9327433596398c255e45a3ecdf70d19a7ba60f23337

Aliases

arxiv: 2605.14117 · arxiv_version: 2605.14117v1 · doi: 10.48550/arxiv.2605.14117 · pith_short_12: WTB52ZRBMY3Z · pith_short_16: WTB52ZRBMY3ZABR4 · pith_short_8: WTB52ZRB

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/WTB52ZRBMY3ZABR43EZHIM2ZMO \
  | 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: b4c3dd6621663790063cd9327433596398c255e45a3ecdf70d19a7ba60f23337

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "41f82cd9c756bc9a80d2f669439e4d1a14438dcbdc180145e816a33b2bc2636b",
    "cross_cats_sorted": [
      "cs.AI"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2026-05-13T21:06:12Z",
    "title_canon_sha256": "d81df676a297d8a4be1bb03038b110a20ce3989b64b91dc74cb62766389197ee"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.14117",
    "kind": "arxiv",
    "version": 1
  }
}