SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents

• URL: http://arxiv.org/abs/2608.29575v1
• Date: Sun, 30 Aug 2026 05:41:20 GMT
• Title: SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents
• Abstract summary: SemTrace is a source-grounded semantic watermark for detecting whether a generated review was influenced by a protected manuscript copy.<n>A protected PDF invisibly carries a content contract that selects one fact from each binary pair and asks an instruction-following reviewer to express those facts in fixed review slots.<n>A frozen natural language inference model then decodes the resulting semantic evidence with explicit erasures and scores the recovered bits against the codeword assigned to that copy.
• Score: 33.910177895525095
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models are increasingly used to read documents and produce downstream text, creating a provenance problem when the document owner cannot control or inspect the model that performs the generation. We introduce SemTrace, a source-grounded semantic watermark for detecting whether a generated review was influenced by a known protected manuscript copy. Rather than biasing token probabilities or imposing surface-form patterns, SemTrace constructs a document-specific binary signature from factual propositions that are directly supported by the manuscript itself. A protected PDF invisibly carries a content contract that selects one fact from each binary pair and asks an instruction-following reviewer to express those facts in fixed review slots without changing its independent evaluation. A frozen natural language inference model then decodes the resulting semantic evidence with explicit erasures and scores the recovered bits against the codeword assigned to that copy. This design targets model-agnostic, assigned-copy exposure detection while keeping the watermark semantically tied to the source document.

Related papers

• WoE Wrote It? Watermarking Mixture-of-Experts LLMs for Black-Box Text Provenance [14.891975420982504]
  We introduce Watermarking of Experts (WoE), a novel black-box text provenance method.<n>WoE biases the vocabulary of specific experts and shifts the watermark signal away from unenforceable inference wrappers.<n>We evaluate WoE across eight MoE models, demonstrating successful watermark detection from suspect text.
  arXiv  Detail & Related papers  (2026-08-29T08:57:06Z)
• Tracing Provenance and Detecting Tampering with Complementary LLM Watermarks [68.64050157343334]
  Existing watermarks preserve provenance under editing, but this same robustness allows an adversary to alter critical content while retaining attribution.<n>We introduce an innovative watermark that jointly provides provenance and tamper evidence.<n>The signals share the same mechanism but use independent keys and different seeding windows over normalized text, making one resilient to edits and the other sensitive to reader-visible changes.
  arXiv  Detail & Related papers  (2026-08-13T01:55:16Z)
• Selective Disclosure Watermarking for Large Language Models [16.579884795891754]
  We propose Hierarchical Vocabulary Routing (HeRo), a watermarking framework that enables selective disclosure of embedded metadata.<n>We show that the proposed scheme preserves the unbiasedness of the underlying sampling process and thus maintains text quality.
  arXiv  Detail & Related papers  (2026-07-06T17:32:01Z)
• SWAN: Semantic Watermarking with Abstract Meaning Representation [61.15962841972814]
  SWAN (Semantic Watermarking with Abstract Meaning Representation) is a novel framework that embeds watermark signatures into the semantic structure of a sentence.<n>As the signature is encoded at the semantic structure level, any paraphrase that preserves meaning automatically preserves the signature.
  arXiv  Detail & Related papers  (2026-05-05T21:13:30Z)
• Unforgeable Watermarks for Language Models via Robust Signatures [12.643204293013007]
  We introduce two novel guarantees: unforgeability and recoverability.<n>We construct the first undetectable watermarking scheme that is robust, unforgeable, and recoverable.
  arXiv  Detail & Related papers  (2026-02-17T03:09:06Z)
• MerkleSpeech: Public-Key Verifiable, Chunk-Localised Speech Provenance via Perceptual Fingerprints and Merkle Commitments [0.0]
  We propose MerkleSpeech, a system for public-key verifiable, chunk-localised speech provenance.<n>The system computes perceptual fingerprints over short speech chunks, commits them in a Merkle tree whose root is signed with an issuer key.<n>We present experiments targeting very low false positive rates under resampling, bandpass filtering, and additive noise.
  arXiv  Detail & Related papers  (2026-02-10T11:58:19Z)
• SWAP: Towards Copyright Auditing of Soft Prompts via Sequential Watermarking [58.475471437150674]
  We propose sequential watermarking for soft prompts (SWAP)<n>SWAP encodes watermarks through a specific order of defender-specified out-of-distribution classes.<n>Experiments on 11 datasets demonstrate SWAP's effectiveness, harmlessness, and robustness against potential adaptive attacks.
  arXiv  Detail & Related papers  (2025-11-05T13:48:48Z)
• Protecting De-identified Documents from Search-based Linkage Attacks [4.065664902103571]
  This paper presents a method to counter search-based linkage attacks while preserving the semantic integrity of the text.<n> Experimental results on a collection of court cases show that the method is able to effectively prevent search-based linkages while remaining faithful to the original content.
  arXiv  Detail & Related papers  (2025-10-07T19:02:21Z)
• Unsupervised Pretraining for Fact Verification by Language Model Distillation [4.504050940874427]
  We propose SFAVEL (Self-supervised Fact Verification via Language Model Distillation), a novel unsupervised pretraining framework. It distils self-supervised features into high-quality claim-fact alignments without the need for annotations. This is enabled by a novel contrastive loss function that encourages features to attain high-quality claim and evidence alignments.
  arXiv  Detail & Related papers  (2023-09-28T15:53:44Z)
• Tracing Text Provenance via Context-Aware Lexical Substitution [81.49359106648735]
  We propose a natural language watermarking scheme based on context-aware lexical substitution. Under both objective and subjective metrics, our watermarking scheme can well preserve the semantic integrity of original sentences.
  arXiv  Detail & Related papers  (2021-12-15T04:27:33Z)
• A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation [50.55448707570669]
  We propose a novel token-level, reference-free hallucination detection task and an associated annotated dataset named HaDes. To create this dataset, we first perturb a large number of text segments extracted from English language Wikipedia, and then verify these with crowd-sourced annotations.
  arXiv  Detail & Related papers  (2021-04-18T04:09:48Z)
• Evidence-Aware Inferential Text Generation with Vector Quantised Variational AutoEncoder [104.25716317141321]
  We propose an approach that automatically finds evidence for an event from a large text corpus, and leverages the evidence to guide the generation of inferential texts. Our approach provides state-of-the-art performance on both Event2Mind and ATOMIC datasets.
  arXiv  Detail & Related papers  (2020-06-15T02:59:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.