Pith Number

pith:O7XK7YSG

pith:2026:O7XK7YSGNACZH7HHYEM2EGJG2T

not attested not anchored not stored refs resolved

LIFT: Last-Mile Fine-Tuning for Table Explicitation

Ashish Tiwari, Divij Khaitan

Last-mile fine-tuning pairs a pre-trained LLM for initial table extraction with a fine-tuned SLM that repairs errors, matching end-to-end SLM fine-tuning on TEDS while using as few as 1,000 examples.

arxiv:2605.13424 v1 · 2026-05-13 · cs.LG · cs.CL

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{O7XK7YSGNACZH7HHYEM2EGJG2T}

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

On a benchmark of 2,596 tables from three datasets, Lift matches or exceeds end-to-end SLM fine-tuning on tree-edit-distance-based similarity (TEDS) metric while requiring as little as 1,000 training examples - where it outperforms end-to-end fine-tuning by up to 0.144 TEDS points.

C2weakest assumption

That errors produced by the pre-trained LLM's initial extraction are consistently repairable by the fine-tuned SLM in a manner that generalizes across the three datasets and varying input formats without the repair step introducing new systematic errors.

C3one line summary

LIFT pairs a pre-trained LLM for initial table extraction with a fine-tuned SLM for error repair, matching end-to-end SLM fine-tuning on TEDS while needing only 1,000 examples and gaining robustness.

References

36 extracted · 36 resolved · 3 Pith anchors

[1] GriTS: Grid Table Similarity Metric for Table Structure Recognition 2023

[2] Image-Based Table Recognition: Data, Model, and Evaluation 2020

[3] Optimized Table Tokenization for Table Structure Recognition 2023

[4] Complicated Table Structure Recognition , author=. 2019 , eprint= 2019

[5] Text-to-Table: A New Way of Information Extraction 2022 · doi:10.18653/v1/2022.acl-long.180

Receipt and verification

First computed	2026-05-18T02:44:47.293217Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

77eeafe246680593fce7c119a21926d4d318943106d4d2216f3f343b6ae22b10

Aliases

arxiv: 2605.13424 · arxiv_version: 2605.13424v1 · doi: 10.48550/arxiv.2605.13424 · pith_short_12: O7XK7YSGNACZ · pith_short_16: O7XK7YSGNACZH7HH · pith_short_8: O7XK7YSG

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/O7XK7YSGNACZH7HHYEM2EGJG2T \
  | 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: 77eeafe246680593fce7c119a21926d4d318943106d4d2216f3f343b6ae22b10

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "18b09cfddbe37d2dc42943e4dcfde90a645396175a2a453fde7ff52e452b24a4",
    "cross_cats_sorted": [
      "cs.CL"
    ],
    "license": "http://creativecommons.org/licenses/by-sa/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-13T12:19:01Z",
    "title_canon_sha256": "5a9e26fac7ebe8ee458f1a225d4223adbfffb5be915fc0730622d3fb656cecbe"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.13424",
    "kind": "arxiv",
    "version": 1
  }
}