Pith Number

pith:SYBJDHAI

pith:2026:SYBJDHAI36UQ5LGRAAHY764TTX

not attested not anchored not stored refs resolved

NEST: Nested Event Stream Transformer for Sequences of Multisets

Benjamin Goldstein, Haoyu Gong, Jillian Hurst, Matthew Engelhard, Minghui Sun, Xingyu You

Preserving the original hierarchy of event streams as sequences of multisets improves both computational efficiency and representation quality in foundation models.

arxiv:2602.00520 v3 · 2026-01-31 · cs.LG

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{SYBJDHAI36UQ5LGRAAHY764TTX}

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

preserving the original hierarchy in the FM architecture provides a useful inductive bias that improves both computational efficiency and representation quality.

C2weakest assumption

That flattening the hierarchy into a one-dimensional sequence necessarily creates spurious within-set relationships and that the original multiset structure supplies a reliable inductive bias.

C3one line summary

NEST is a nested transformer for sequences of multisets that uses masked set modeling to learn improved set-level representations from hierarchical event streams like EHRs.

References

42 extracted · 42 resolved · 3 Pith anchors

[1] Longformer: The Long-Document Transformer 2004 · arXiv:2004.05150

[2] Neural legal judgment prediction in english 2019

[3] An exploration of hierarchical attention transformers for efficient long document classification.arXiv preprint arXiv:2210.05529, 2022 2022

[4] Diffcse: Difference-based contrastive learning for sentence embeddings 2022

[5] Bert: Pre-training of deep bidi- rectional transformers for language understanding 2019

Receipt and verification

First computed	2026-05-17T23:39:16.483199Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

9602919c08dfa90eacd1000f8ffb939de37abd5e37724a5b5b0eb295d9112f56

Aliases

arxiv: 2602.00520 · arxiv_version: 2602.00520v3 · doi: 10.48550/arxiv.2602.00520 · pith_short_12: SYBJDHAI36UQ · pith_short_16: SYBJDHAI36UQ5LGR · pith_short_8: SYBJDHAI

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/SYBJDHAI36UQ5LGRAAHY764TTX \
  | 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: 9602919c08dfa90eacd1000f8ffb939de37abd5e37724a5b5b0eb295d9112f56

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "e9262779e95c0b42d82984f39f1cfc416ace514913b0ae6c3bacefbeaa30a479",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-01-31T05:21:27Z",
    "title_canon_sha256": "bbc190419284b0fd364101cfa0d42c85d71fc8d28e5bda0f21b376872e54b45e"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2602.00520",
    "kind": "arxiv",
    "version": 3
  }
}