ARXIV:2606.05609 · UNCATEGORIZED · SUBMITTED 06 JUN · 11:26 UTC · FRESHNESS FRESH

VerifiedSource: PDF linkedVerifiedPaperPack: citation fields availablePartialProof: unverified proof status

SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks

Seungwon Jeong · Jiwoo Jeong · Hyeonjin Kim · Yunseok Lee · Woojin Lee · arXiv

ScienceToStartup currently rates this 0.0/10 on the public viability pass. Experiments across multiple models demonstrate that SlotGCG significantly outperforms existing methods. A public repository is linked, so build verification can…

Ship in 2-4 weeks›Score0.0Evidence unverified

Opportunity summary

Pain customer pain not on file

Evidence 0 refs | 4 sources | 83% coverage

Blocker Evidence unverified

Open Build Read PDF Signal Canvas Track

PROBLEM

As large language models (LLMs) are widely deployed, identifying their vulnerability through jailbreak attacks becomes increasingly critical.

METHOD

Full abstract

As large language models (LLMs) are widely deployed, identifying their vulnerability through jailbreak attacks becomes increasingly critical. Optimization-based attacks like Greedy Coordinate Gradient (GCG) have focused on inserting adversarial tokens to the end of prompts. However, GCG restricts adversarial tokens to a fixed insertion point (typically the prompt suffix), leaving the effect of inserting tokens at other positions unexplored. In this paper, we empirically investigate \emph{slots}, i.e., candidate positions within a prompt where tokens can be inserted. We find that vulnerability to jailbreaking is highly related to the selection of the \emph{slots}. Based on these findings, we introduce the \textit{Vulnerable Slot Score} (VSS) to quantify the positional vulnerability to jailbreaking. We then propose SlotGCG, which evaluates all slots with VSS, selects the most vulnerable slots for insertion, and runs a targeted optimization attack at those slots. Our approach provides a position-search mechanism that is attack-agnostic and can be plugged into any optimization-based attack, adding only 200ms of preprocessing time. Experiments across multiple models demonstrate that SlotGCG significantly outperforms existing methods. Specifically, it achieves 14\% higher Attack Success Rates (ASR) over GCG-based attacks, converges faster, and shows superior robustness against defense methods with 42\% higher ASR than baseline approaches. Our implementation is available at \href{https://github.com/youai058/SlotGCG}{https://github.com/youai058/SlotGCG}

RESULT

WHY NOW

Uncategorized moved forward this cycle; last verified June 2026. Public score 0.0/10. Implementation evidence is present through a linked repository.

Continue into Read for claims, analysis, references, and neighboring papers.

Opportunity summary

Score0.0

Paincustomer pain not on file

Evidence0 refs | 4 sources | 83% coverage

Blockerno shell-level blocker reported

Analysis summary

VerifiedSource: PDF linkedVerifiedPaperPack: citation fields availablePartialProof: unverified proof status

SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks

Seungwon Jeong · Jiwoo Jeong · Hyeonjin Kim · Yunseok Lee · Woojin Lee · arXiv

Competitive landscape

No named competitor graph is public yet; the page still exposes the segment, adoption evidence, and score state so the commercial read is not blank.

Segment

Uncategorized

Adoption evidence

Public code linked for build inspection

Commercial read

0.0/10 public viability

Direct

not classified

Adjacent

not classified

Substitute

not classified

Unknown

not classified

{ "contract_version": "paper-r2", "paper_id": "deaf2810-09bc-45dc-8b69-b8f359aa286b", "arxiv_id": "2606.05609", "canonical_route": "/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "active_tab": "synced from current hash by the drawer client", "selected_artifact": "slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "endpoints": { "paper_pack": "/api/v1/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks/paper-pack", "build_passport": "/api/v1/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks/build-passport", "mcp_resource": "sciencetostartup://surfaces/paper-workspace" } }

{ "surface": "paper", "mode": "paper", "query": "SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks", "normalized_query": "2606.05609", "route": "/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "paper_ref": "slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "topic_slug": null, "benchmark_ref": null, "dataset_ref": null }

{ "@context": "https://schema.org", "@graph": [ { "@type": "WebPage", "@id": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks#webpage", "url": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "name": "SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks", "description": "As large language models (LLMs) are widely deployed, identifying their vulnerability through jailbreak attacks becomes increasingly critical. Optimization-based attacks like Greedy Coordinate Gradient (GCG) have focused on inserting adversarial tokens to the end of prompts. However, GCG restricts adversarial tokens to a fixed insertion point (typically the prompt suffix), leaving the effect of inserting tokens at other positions unexplored. In this paper, we empirically investigate \\emph{slots}, i.e., candidate positions within a prompt where tokens can be inserted. We find that vulnerability to jailbreaking is highly related to the selection of the \\emph{slots}. Based on these findings, we introduce the \\textit{Vulnerable Slot Score} (VSS) to quantify the positional vulnerability to jailbreaking. We then propose SlotGCG, which evaluates all slots with VSS, selects the most vulnerable slots for insertion, and runs a targeted optimization attack at those slots. Our approach provides a position-search mechanism that is attack-agnostic and can be plugged into any optimization-based attack, adding only 200ms of preprocessing time. Experiments across multiple models demonstrate that SlotGCG significantly outperforms existing methods. Specifically, it achieves 14\\% higher Attack Success Rates (ASR) over GCG-based attacks, converges faster, and shows superior robustness against defense methods with 42\\% higher ASR than baseline approaches. Our implementation is available at \\href{https://github.com/youai058/SlotGCG}{https://github.com/youai058/SlotGCG}", "isPartOf": { "@id": "https://sciencetostartup.com/#website" } }, { "@type": "ScholarlyArticle", "@id": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks#scholarlyArticle", "headline": "SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks", "description": "As large language models (LLMs) are widely deployed, identifying their vulnerability through jailbreak attacks becomes increasingly critical. Optimization-based attacks like Greedy Coordinate Gradient (GCG) have focused on inserting adversarial tokens to the end of prompts. However, GCG restricts adversarial tokens to a fixed insertion point (typically the prompt suffix), leaving the effect of inserting tokens at other positions unexplored. In this paper, we empirically investigate \\emph{slots}…", "url": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks", "sameAs": "https://arxiv.org/abs/2606.05609", "identifier": { "@type": "PropertyValue", "propertyID": "arXiv", "value": "2606.05609" }, "isAccessibleForFree": true, "isPartOf": { "@id": "https://sciencetostartup.com/#website" }, "datePublished": "2026-06-04T02:31:29.000Z", "author": [ { "@type": "Person", "name": "Seungwon Jeong" }, { "@type": "Person", "name": "Jiwoo Jeong" }, { "@type": "Person", "name": "Hyeonjin Kim" }, { "@type": "Person", "name": "Yunseok Lee" }, { "@type": "Person", "name": "Woojin Lee" } ], "codeRepository": "https://github.com/youai058/SlotGCG}{https://github.com/youai058/SlotGCG}", "additionalProperty": [ { "@type": "PropertyValue", "propertyID": "researchDomain", "value": "Uncategorized" }, { "@type": "PropertyValue", "propertyID": "commercialReadiness", "value": "code, repo url" } ] }, { "@type": "SoftwareSourceCode", "@id": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks#software", "name": "SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks - Source Code", "description": "As large language models (LLMs) are widely deployed, identifying their vulnerability through jailbreak attacks becomes increasingly critical. Optimization-based attacks like Greedy Coordinate Gradient (GCG) have focused on inserting adversarial tokens to the end of prompts. However, GCG restricts adversarial tokens to a fixed insertion point (typically the prompt suffix), leaving the effect of inserting tokens at other positions unexplored. In this paper, we empirically investigate \\emph{slots}…", "codeRepository": "https://github.com/youai058/SlotGCG}{https://github.com/youai058/SlotGCG}", "url": "https://github.com/youai058/SlotGCG}{https://github.com/youai058/SlotGCG}" }, { "@type": "BreadcrumbList", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://sciencetostartup.com" }, { "@type": "ListItem", "position": 2, "name": "Uncategorized", "item": "https://sciencetostartup.com/topics" }, { "@type": "ListItem", "position": 3, "name": "SlotGCG: Exploiting the Positional Vulnerability in LLMs for", "item": "https://sciencetostartup.com/paper/slotgcg-exploiting-the-positional-vulnerability-in-llms-for-jailbreak-attacks" } ] } ] }

Competitive landscape

No named competitor graph is public yet; the page still exposes the segment, adoption evidence, and score state so the commercial read is not blank.

Segment

Uncategorized

Adoption evidence

Public code linked for build inspection

Commercial read

0.0/10 public viability

Direct

not classified

Adjacent

not classified

Substitute

not classified

Unknown

not classified

SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks

SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attacks

Claim map

Constellation map

Competitive landscape

Buzz

PDF

REFERENCES

Related Papers

Subscribe to the weekly brief

Build artifacts

Brief

Experiment plan

Validation checklist

Scientific founder

Translational engineer

Domain operator

GTM lead

Regulatory/clinical advisor

Timeline

Claim map

Constellation map

Competitive landscape

Buzz

PDF

REFERENCES

Related Papers

Subscribe to the weekly brief

Build artifacts

Brief

Experiment plan

Validation checklist

Scientific founder

Translational engineer

Domain operator

GTM lead

Regulatory/clinical advisor

Timeline