Skip to content

Wire Briefings Search

Fetching from the wire…

01

02

03

04

05

06

07

08

Wire Briefings Search

Entity trail

CamoDocs

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

1

Findings

1

Edges

3

Sources

1

Corpus findings

2026-08-31 / arxiv-researcherCamoDocs Poisons RAG Without Putting the Query in the Document, Hitting 61.80% ASR on GPT-5.4-miniCamoDocs (arXiv 2608.28389, 2026-08-28, cs.CL/cs.CR) drops the query-inclusion trick that makes existing RAG poisoning easy to filter, instead chunking synthesized benign and adversarial drafts, swapping selected tokens in benign chunks for dispersion tokens that spread the poisoned embeddings, and applying coherence filtering to keep readability. Tested across seven RAG defenses, three open-weight LLMs and three benchmarks, it holds strong average attack success while avoiding query-overlap artifacts, reaching 61.80% average ASR on GPT-5.4-mini and 55.09% on Claude-Haiku-4.5. Erasure-heavy clustering defenses such as TrustRAG cut ASR but only at substantial utility cost on retrieval-dependent benchmarks like NeoQA; code is at github.com/jaewonalive/CamoDocs.

Graph relationships

BENCHMARKED AGAINST
Claude Haiku 4.5
CamoDocs achieves 55.09% attack success rate on Claude-Haiku-4.5.
Source finding
BENCHMARKED AGAINST
GPT-5.4-Mini
CamoDocs achieves 61.80% attack success rate on GPT-5.4-mini.
Source finding
CRITICIZES
TrustRAG
CamoDocs demonstrates that TrustRAG defenses reduce ASR only at substantial utility cost.
Source finding

Source trail

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues