Pith Number

pith:BD6QRVPD

pith:2026:BD6QRVPDQMXYDLIYGLHPGSIEQ5

not attested not anchored not stored refs resolved

Privacy Evaluation of Generative Models for Trajectory Generation

Chiara Pugliese, Chiara Renso, Emanuele Carlini, Francesco Lettich, Hanna Kavalionak, Ioannis Kontopoulos, Konstantinos Tserpes, Stavros Bouras

Generative models do not guarantee privacy for synthetic trajectory data

arxiv:2605.15246 v1 · 2026-05-14 · 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{BD6QRVPDQMXYDLIYGLHPGSIEQ5}

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

Although these models are often assumed to preserve privacy due to their generative nature, this assumption does not necessarily hold, as demonstrated by the feasibility of using Membership Inference Attacks against representative models.

C2weakest assumption

That standard membership inference attack methods are directly applicable to trajectory generation tasks and that success on the chosen representative models indicates general privacy leakage risks across generative trajectory approaches.

C3one line summary

Generative models for trajectory data do not inherently preserve privacy, as membership inference attacks can identify training data points in representative models.

References

48 extracted · 48 resolved · 3 Pith anchors

[1] Unique in the crowd: The privacy bounds of human mobility, 2013

[2] Trajectory generation: a survey on methods and techniques, 2025

[3] The secret sharer: Evaluating and testing unintended memorization in neural net- works, 2019

[4] Sok: Can trajectory generation combine privacy and utility? 2024

[5] Mobility trajectory generation: a survey, 2023

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-20T00:00:48.281378Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

08fd08d5e3832f81ad1832cef349048779cbb18dee499062c07cf6886c2a2f37

Aliases

arxiv: 2605.15246 · arxiv_version: 2605.15246v1 · doi: 10.48550/arxiv.2605.15246 · pith_short_12: BD6QRVPDQMXY · pith_short_16: BD6QRVPDQMXYDLIY · pith_short_8: BD6QRVPD

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/BD6QRVPDQMXYDLIYGLHPGSIEQ5 \
  | 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: 08fd08d5e3832f81ad1832cef349048779cbb18dee499062c07cf6886c2a2f37

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "eba8f1801d42538c52a3d4f015a4f69eee6e60c6011e6199dbd1cb0131d1a19d",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-14T10:57:34Z",
    "title_canon_sha256": "b4b6075aa543763a76f44a984f8635c4f6a9b0e5aa98a504f73fc6f0b6366858"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15246",
    "kind": "arxiv",
    "version": 1
  }
}