Paper Citation Record · LEDGER

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

As of 24 August 2026, this Paper Citation Record lists 100 of 115 outbound references and 6 inbound Pith citation observations for arXiv:2501.15269.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2501.15269 v1

Coverage vector

measured 100 of 115 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-10T14:29:28.341757Z

measured 106 of 106 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-23T06:30:58.430688+00:00

measured 6 of 6 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-06T15:32:34.540866Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: arxiv_reference, observed 2026-06-29T22:34:01.805164Z

Reference resolution

100 of 115 outbound references displayed

verified exact3
verified fuzzy12
unresolved85
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 1085029e-f5bc-4e3e-b53f-e9c6969a3217 · outbound

This paper cites GPT-4 Technical Report.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink GPT-4 Technical Report

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.846865Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 801c4a88-2033-41f8-81f3-88ed53b86144 · outbound

This paper cites Flamingo: a visual language model for few-shot learning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Flamingo: a visual language model for few-shot learning

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.851227Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 85acb3bb-d89c-4fd5-b14f-e11da302d963 · outbound

This paper cites Vqa: Visual question answering.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Vqa: Visual question answering

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.854969Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d7d1d599-7e2e-416e-9651-168a7cef0606 · outbound

This paper cites Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.858635Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8b541e01-9d04-44b2-b5fd-2717b63d2b97 · outbound

This paper cites Hallucination of Multimodal Large Language Models: A Survey.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Hallucination of Multimodal Large Language Models: A Survey

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.863635Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 41ac2492-6092-4054-903f-a45f05b2f410 · outbound

This paper cites The (r) evolution of multimodal large language models: A survey.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink The (r) evolution of multimodal large language models: A survey

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.867663Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1d59d971-b065-444d-ad33-f710b95db610 · outbound

This paper cites Alleviating Hallucinations in Large Vision-Language Models through Hallucination-Induced Optimization.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Alleviating Hallucinations in Large Vision-Language Models through Hallucination-Induced Optimization

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.871553Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dd212740-b38c-48ab-be8d-5b4d29ca6468 · outbound

This paper cites Lion: Empowering multimodal large language model with dual-level visual knowledge.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Lion: Empowering multimodal large language model with dual-level visual knowledge

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.876564Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a308abfb-da00-43d2-b970-4cca791c8646 · outbound

This paper cites ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.880204Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 847f7306-1e7e-4928-b61e-3f490958965d · outbound

This paper cites MiniGPT-v2: large language model as a unified interface for vision-language multi-task learning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink MiniGPT-v2: large language model as a unified interface for vision-language multi-task learning

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.884637Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 189f1814-aa29-4386-aced-6b0188ad64aa · outbound

This paper cites IPL: Leveraging Multimodal Large Language Models for Intelligent Product Listing.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink IPL: Leveraging Multimodal Large Language Models for Intelligent Product Listing

Reference 11

Resolution: verified exact; local_arxiv, observed 2026-08-10T14:29:28.976575Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 74dc016f-f815-49f1-bb9b-f7ef715dfa01 · outbound

This paper cites Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.892821Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation db0fafe1-74d4-4dfa-9aba-70883dead080 · outbound

This paper cites ShareGPT4V: Improving Large Multi-Modal Models with Better Captions.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink ShareGPT4V: Improving Large Multi-Modal Models with Better Captions

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.896365Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 74352b39-8c25-4ace-887d-21071d6d40a5 · outbound

This paper cites Multi-Object Hallucination in Vision-Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Multi-Object Hallucination in Vision-Language Models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.900835Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5052bb9d-86ae-4270-ba46-a70f7f0b920b · outbound

This paper cites Transfer Attack for Bad and Good: Explain and Boost Adversarial Transferability across Multimodal Large Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Transfer Attack for Bad and Good: Explain and Boost Adversarial Transferability across Multimodal Large Language Models

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.904771Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9cd4ca39-9df0-4cff-8292-9f30898ba5d7 · outbound

This paper cites Vicuna: An open-source chatbot impressing gpt-4 with 90%* chatgpt quality.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Vicuna: An open-source chatbot impressing gpt-4 with 90%* chatgpt quality

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.908486Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2123aedd-9f85-4e36-ad58-fd53aebdc76a · outbound

This paper cites Optimal structure identifi- cation with greedy search.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Optimal structure identifi- cation with greedy search

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.911856Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 552ab65e-a6f6-4d2a-bd3e-0cf30613f362 · outbound

This paper cites On the robustness of large multimodal models against image adversarial attacks.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink On the robustness of large multimodal models against image adversarial attacks

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.915532Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0bf73e6b-7975-4970-80d8-ef46c0e2d9c6 · outbound

This paper cites InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.919151Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f2de51ba-2160-4adf-8458-b05c90737a74 · outbound

This paper cites Vision transformers need registers.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Vision transformers need registers

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.923851Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f2f4adb4-faaf-4e59-8bb2-bbc09de2f695 · outbound

This paper cites HiLM-D: Enhancing MLLMs with Multi-Scale High-Resolution Details for Autonomous Driving.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink HiLM-D: Enhancing MLLMs with Multi-Scale High-Resolution Details for Autonomous Driving

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.927310Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 58028e65-4650-4515-a6d0-bc4e182c951c · outbound

This paper cites An image is worth 16x16 words: Transformers for image recognition at scale.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink An image is worth 16x16 words: Transformers for image recognition at scale

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.931147Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f1ef4f17-242d-4549-84c6-fc455f335913 · outbound

This paper cites Eva: Exploring the limits of masked visual representation learning at scale.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Eva: Exploring the limits of masked visual representation learning at scale

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.935698Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5505cf1a-0d08-4870-bca4-566f3218cabc · outbound

This paper cites MME-Finance: A Multimodal Finance Benchmark for Expert-level Understanding and Reasoning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink MME-Finance: A Multimodal Finance Benchmark for Expert-level Understanding and Reasoning

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.939969Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 52052379-10f3-4a09-a9f3-4e07a1fdd88d · outbound

This paper cites Adversarial robustness for visual ground- ing of multimodal large language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Adversarial robustness for visual ground- ing of multimodal large language models

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.943665Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6e8dad2c-b584-4946-840e-5eea5a6ab8c8 · outbound

This paper cites Inducing high energy-latency of large vision-language models with verbose images.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Inducing high energy-latency of large vision-language models with verbose images

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.947472Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a6e18ee9-723c-42ce-ae64-2d9d3a993120 · outbound

This paper cites MultiModal-GPT: A Vision and Language Model for Dialogue with Humans.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink MultiModal-GPT: A Vision and Language Model for Dialogue with Humans

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.951706Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e8daa2fc-b5c4-433c-9815-f0bc5301c395 · outbound

This paper cites FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.955780Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bd183a0c-0e20-469a-b8e4-8ae61fac2a91 · outbound

This paper cites Hallusionbench: an advanced diagnostic suite for entangled language hal- lucination and visual illusion in large vision-language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Hallusionbench: an advanced diagnostic suite for entangled language hal- lucination and visual illusion in large vision-language models

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.959655Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c998b18e-4e50-4bca-b5b2-1741d40c3e0a · outbound

This paper cites Skip \n: A Simple Method to Reduce Hallucination in Large Vision-Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Skip \n: A Simple Method to Reduce Hallucination in Large Vision-Language Models

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.963407Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b7430202-b296-46fe-bf57-0c57a47a80db · outbound

This paper cites Clipscore: A reference-free 16 evaluation metric for image captioning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Clipscore: A reference-free 16 evaluation metric for image captioning

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.968378Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cc13ba67-40da-402c-a6d0-57f9316487a2 · outbound

This paper cites The Curious Case of Neural Text Degeneration.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink The Curious Case of Neural Text Degeneration

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.972140Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3cda6207-2bab-46be-b681-479e0fc498f9 · outbound

This paper cites Naturalistic physical adversarial patch for object detectors.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Naturalistic physical adversarial patch for object detectors

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.975776Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 20466ca6-7dd4-417d-8887-b9d8e8b146ee · outbound

This paper cites Opera: Alleviating hallucination in multi-modal large language models via over-trust penalty and retrospection-allocation.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Opera: Alleviating hallucination in multi-modal large language models via over-trust penalty and retrospection-allocation

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.979180Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3acce825-9769-4724-8fa4-3b7a1478bb33 · outbound

This paper cites Hallucination augmented contrastive learning for multimodal large language model.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Hallucination augmented contrastive learning for multimodal large language model

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.982581Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 76070ec9-ee0f-4401-a658-c7b443852965 · outbound

This paper cites Diffattack: Evasion attacks against diffusion-based adversarial pu- rification.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Diffattack: Evasion attacks against diffusion-based adversarial pu- rification

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.985981Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4a662d7d-fbcd-4f5a-b8eb-fe050e5cb1d0 · outbound

This paper cites Vi- sual genome: Connecting language and vision using crowdsourced dense image annotations.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Vi- sual genome: Connecting language and vision using crowdsourced dense image annotations

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.989642Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3fb5a997-1bb6-4129-8c84-3828e00c09ac · outbound

This paper cites Gemini Pro Defeated by GPT-4V: Evidence from Education.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Gemini Pro Defeated by GPT-4V: Evidence from Education

Reference 38

Resolution: verified exact; local_arxiv, observed 2026-08-10T14:29:28.844132Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 8971a31e-e55d-48d4-ae24-075e0739975a · outbound

This paper cites Robust evaluation of diffusion-based adversarial purification.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Robust evaluation of diffusion-based adversarial purification

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:27.996615Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0b5459d7-e655-4f96-b2ef-3fe2e08bca7e · outbound

This paper cites Volcano: Mitigating Multimodal Hallucination through Self-Feedback Guided Revision.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Volcano: Mitigating Multimodal Hallucination through Self-Feedback Guided Revision

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.000337Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8f18a498-90f4-4086-a615-a9f29c64188a · outbound

This paper cites Miti- gating object hallucinations in large vision-language models through visual contrastive decoding.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Miti- gating object hallucinations in large vision-language models through visual contrastive decoding

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.004970Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3c96ac91-d802-4803-88e7-5b8a7db3f860 · outbound

This paper cites Llava-med: Training a large language-and-vision assistant for biomedicine in one day.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Llava-med: Training a large language-and-vision assistant for biomedicine in one day

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.008354Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 31850786-e223-45f7-b996-6f84eb60e2d2 · outbound

This paper cites Manipllm: Embodied multimodal large language model for object-centric robotic manipula- tion.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Manipllm: Embodied multimodal large language model for object-centric robotic manipula- tion

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.012256Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 48082089-62de-4957-87dd-63aa377741fb · outbound

This paper cites Evaluating Object Hallucination in Large Vision-Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Evaluating Object Hallucination in Large Vision-Language Models

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.017232Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3dd82601-0890-4716-a2cc-6dbf097ef0be · outbound

This paper cites Monkey: Image resolution and text label are important things for large multi-modal models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Monkey: Image resolution and text label are important things for large multi-modal models

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.021367Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ae855e47-b807-4cac-9d5d-c50bb6ff6c3a · outbound

This paper cites Harnessing GPT-4V(ision) for Insurance: A Preliminary Exploration.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Harnessing GPT-4V(ision) for Insurance: A Preliminary Exploration

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.025287Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e30fc5b9-0107-450e-979a-552c7b088466 · outbound

This paper cites Microsoft coco: Common objects in context.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Microsoft coco: Common objects in context

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.029799Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 790a7b4f-77f9-48d3-9e00-37211a188720 · outbound

This paper cites Interpreting and Mitigating Hallucination in MLLMs through Multi-agent Debate.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Interpreting and Mitigating Hallucination in MLLMs through Multi-agent Debate

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.033866Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9e3a6a25-5c4f-4a5b-bd2e-d60400782eb9 · outbound

This paper cites GEMeX: A Large-Scale, Groundable, and Explainable Medical VQA Benchmark for Chest X-ray Diagnosis.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink GEMeX: A Large-Scale, Groundable, and Explainable Medical VQA Benchmark for Chest X-ray Diagnosis

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.038592Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2c1a49b8-b6af-4e6d-a559-f4ce8a8a6cab · outbound

This paper cites Mitigating hallucination in large multi-modal models via robust instruction tuning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Mitigating hallucination in large multi-modal models via robust instruction tuning

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.042623Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 33fab8f4-cdec-4733-bad3-f3c3bf1ab5ba · outbound

This paper cites Improved baselines with visual instruction tun- ing.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Improved baselines with visual instruction tun- ing

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.046459Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6f6c710a-1f2b-413b-97d7-ec4aebaa2cb2 · outbound

This paper cites Visual instruction tuning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Visual instruction tuning

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.049942Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 347c487f-2aa3-4030-b7f3-d5412e5620d4 · outbound

This paper cites Models See Hallucinations: Evaluating the Factuality in Video Captioning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Models See Hallucinations: Evaluating the Factuality in Video Captioning

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.053371Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4a53f1a0-06e6-4818-8908-e0e4d35b268c · outbound

This paper cites PhD: A ChatGPT-Prompted Visual hallucination Evaluation Dataset.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink PhD: A ChatGPT-Prompted Visual hallucination Evaluation Dataset

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.056851Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 32dc33f8-9e26-4b78-869c-fc45190565cc · outbound

This paper cites MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.060424Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 20260262-0ed8-4ab3-8e44-f8ddfd4f377c · outbound

This paper cites MMInstruct: A High-Quality Multi-Modal Instruction Tuning Dataset with Extensive Diversity.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink MMInstruct: A High-Quality Multi-Modal Instruction Tuning Dataset with Extensive Diversity

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.064100Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 83840c11-7195-427f-a43d-73e3cacd31fe · outbound

This paper cites Robollm: Robotic vision tasks grounded on multimodal large language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Robollm: Robotic vision tasks grounded on multimodal large language models

Reference 57

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.067564Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8dd12174-451e-401a-8b1d-90ec30f71686 · outbound

This paper cites JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.071084Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d06010ab-dcd7-464f-9472-a6daabaff6c6 · outbound

This paper cites Visual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Image Character.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Visual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Image Character

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.074482Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fb2887a6-1cb3-47c2-b0b9-b24129c5184f · outbound

This paper cites Ok-vqa: A visual question answering benchmark requiring external knowledge.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Ok-vqa: A visual question answering benchmark requiring external knowledge

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.078245Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1c4a6f49-ac79-4291-a5ef-452841b5138d · outbound

This paper cites Large Language Models: A Survey.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Large Language Models: A Survey

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.081919Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6a8bd129-2a24-45c1-b7bc-743a381b7d52 · outbound

This paper cites Yo'LLaVA: Your Personalized Language and Vision Assistant.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Yo'LLaVA: Your Personalized Language and Vision Assistant

Reference 62

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.086662Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d1e7d157-24e3-45a2-af25-c004c017afe1 · outbound

This paper cites Jailbreaking Attack against Multimodal Large Language Model.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Jailbreaking Attack against Multimodal Large Language Model

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.090315Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation adf28c34-7f99-4986-9011-ebd7b34f0dcb · outbound

This paper cites Gpt-4v(ision) technical work and authors.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Gpt-4v(ision) technical work and authors

Reference 64

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.094725Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fb5e4e67-91c0-4dbf-b3e5-1215c72c1928 · outbound

This paper cites Gemini flash.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Gemini flash

Reference 65

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.098994Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4c4e6060-9076-4d9d-821b-387505eb533b · outbound

This paper cites Gpt-4o mini: advancing cost-efficient intelli- gence.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Gpt-4o mini: advancing cost-efficient intelli- gence

Reference 66

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.102622Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0fc84af7-7740-4afa-b1b7-2c771c7b9d6b · outbound

This paper cites Gemini Goes to Med School: Exploring the Capabilities of Multimodal Large Language Models on Medical Challenge Problems & Hallucinations.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Gemini Goes to Med School: Exploring the Capabilities of Multimodal Large Language Models on Medical Challenge Problems & Hallucinations

Reference 67

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.106071Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 34d20abf-4daa-4e04-ac0e-de1bd311da0c · outbound

This paper cites Kosmos-2: Grounding Multimodal Large Language Models to the World.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Kosmos-2: Grounding Multimodal Large Language Models to the World

Reference 68

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.110780Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4cceb0a5-18c8-41fa-8f02-c3161de806cb · outbound

This paper cites Grounding multimodal large language models to the world.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Grounding multimodal large language models to the world

Reference 69

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.114766Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fdd19afd-82a8-4ef0-a335-634424fb29d3 · outbound

This paper cites Visual adversarial examples jailbreak aligned large language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Visual adversarial examples jailbreak aligned large language models

Reference 70

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.118822Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca80f517-2ac1-4712-bd7f-3c234b98d298 · outbound

This paper cites Learning transferable visual models from natural lan- guage supervision.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Learning transferable visual models from natural lan- guage supervision

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.122547Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7b31e1eb-3310-4da8-82b9-6debc1f1dbeb · outbound

This paper cites Object hallucination in image captioning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Object hallucination in image captioning

Reference 72

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.387161Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 7be458b5-51b1-4e2d-8ad1-e8c301c9cee0 · outbound

This paper cites Laion coco: 600m synthetic cap- tions from laion2b-en.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Laion coco: 600m synthetic cap- tions from laion2b-en

Reference 73

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.375809Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 240c93c4-f465-4f6f-a528-098b219eaf5a · outbound

This paper cites Jailbreak in pieces: Compositional adversarial attacks on multi-modal language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Jailbreak in pieces: Compositional adversarial attacks on multi-modal language models

Reference 74

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.364508Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation d9d1ea95-bad3-484c-8f22-4482098919c9 · outbound

This paper cites Prefixing Attention Sinks can Mitigate Activation Outliers for Large Language Model Quantization.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Prefixing Attention Sinks can Mitigate Activation Outliers for Large Language Model Quantization

Reference 75

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.240127Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0d0849ed-ecd4-4787-9f36-e4c290ebaccc · outbound

This paper cites Self-training large language and vision assistant for medical question answering.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Self-training large language and vision assistant for medical question answering

Reference 76

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.354145Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation c4fc7447-6fa3-4e0c-aac4-f959644e3131 · outbound

This paper cites Aligning Large Multimodal Models with Factually Augmented RLHF.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Aligning Large Multimodal Models with Factually Augmented RLHF

Reference 77

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.248761Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e4e75d29-822e-4c62-bd20-0cba3bcc7332 · outbound

This paper cites An Empirical Study and Analysis of Text-to-Image Generation Using Large Language Model-Powered Textual Representation.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink An Empirical Study and Analysis of Text-to-Image Generation Using Large Language Model-Powered Textual Representation

Reference 78

Resolution: verified exact; local_arxiv, observed 2026-08-10T14:29:28.629907Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation d874f987-be29-40bd-a1f7-94cac49a0673 · outbound

This paper cites LLaMA: Open and Efficient Foundation Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink LLaMA: Open and Efficient Foundation Language Models

Reference 79

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.257332Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 92c8cdad-cf58-48c9-b88d-4d0257b2aa59 · outbound

This paper cites Attention Is All You Need.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Attention Is All You Need

Reference 80

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.261514Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 419f1e5e-7dda-4b92-89f3-0dbd140b6289 · outbound

This paper cites AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation

Reference 81

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.265994Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation efb8a9ea-9688-4b1d-894b-bf58ac43da7b · outbound

This paper cites Label words are anchors: An information flow perspective for under- standing in-context learning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Label words are anchors: An information flow perspective for under- standing in-context learning

Reference 82

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.343500Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation d42c0744-2cd9-43fc-bd22-079b58f5946e · outbound

This paper cites AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting

Reference 83

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.274493Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 12669dc7-45b5-4a1c-88bf-31dc496043a2 · outbound

This paper cites Can GPT-4V(ision) Serve Medical Applications? Case Studies on GPT-4V for Multimodal Medical Diagnosis.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Can GPT-4V(ision) Serve Medical Applications? Case Studies on GPT-4V for Multimodal Medical Diagnosis

Reference 84

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.278563Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1c2e545c-89de-487b-815d-27e839f68b4e · outbound

This paper cites Efficient streaming language models with attention sinks.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Efficient streaming language models with attention sinks

Reference 85

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.332619Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 06624e22-3b03-4c6f-8cbd-6adc3c90ae81 · outbound

This paper cites Open-FinLLMs: Open Multimodal Large Language Models for Financial Applications.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Open-FinLLMs: Open Multimodal Large Language Models for Financial Applications

Reference 86

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.285836Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5347bd0a-9cd8-4195-824c-e6edb3403e8c · outbound

This paper cites Self-evaluation guided beam search for reasoning.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Self-evaluation guided beam search for reasoning

Reference 87

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.322101Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation c8f13d1e-e2bb-49cb-8db6-647abbe28d04 · outbound

This paper cites Drivegpt4: Interpretable end-to-end au- tonomous driving via large language model.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Drivegpt4: Interpretable end-to-end au- tonomous driving via large language model

Reference 88

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.310497Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation be00772f-ab01-4e96-9645-29dd506a6267 · outbound

This paper cites UReader: Universal OCR-free Visually-situated Language Understanding with Multimodal Large Language Model.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink UReader: Universal OCR-free Visually-situated Language Understanding with Multimodal Large Language Model

Reference 89

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.296141Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 68a55dc5-5502-4a89-ab29-33ee49b962d3 · outbound

This paper cites mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality

Reference 90

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.299813Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c03b8bc0-1090-43d1-aa14-905524ae4ca9 · outbound

This paper cites Woodpecker: Hallucination Correction for Multimodal Large Language Models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Woodpecker: Hallucination Correction for Multimodal Large Language Models

Reference 91

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.303331Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1133fe17-9355-436b-956f-2df28ff037be · outbound

This paper cites Unveiling the Safety of GPT-4o: An Empirical Study using Jailbreak Attacks.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Unveiling the Safety of GPT-4o: An Empirical Study using Jailbreak Attacks

Reference 92

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.307138Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 532e901e-98eb-49ec-bcb3-da71f6749e5f · outbound

This paper cites Hallucidoctor: Mitigating hallucinatory toxicity in visual instruction data.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Hallucidoctor: Mitigating hallucinatory toxicity in visual instruction data

Reference 93

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.299253Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation ba79c143-3998-4f6a-8444-4a59f9366f26 · outbound

This paper cites Rlhf-v: Towards trustworthy mllms via behavior alignment from fine- grained correctional human feedback.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Rlhf-v: Towards trustworthy mllms via behavior alignment from fine- grained correctional human feedback

Reference 94

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.287588Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation f379fb86-6723-4b09-8d16-9290b406a22e · outbound

This paper cites Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibration.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibration

Reference 95

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.318562Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation db51e8cf-b656-4296-9df9-53d5ca57e3d8 · outbound

This paper cites Less is More: Mitigating Multimodal Hallucination from an EOS Decision Perspective.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Less is More: Mitigating Multimodal Hallucination from an EOS Decision Perspective

Reference 96

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.323931Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f1dfbae8-9876-4e0c-a730-6dd11911cf5b · outbound

This paper cites LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding

Reference 97

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.328190Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c05361fe-9d17-4206-a37d-5467f0ed5f7d · outbound

This paper cites On evaluating adversarial robustness of large vision- language models.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink On evaluating adversarial robustness of large vision- language models

Reference 98

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.274888Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation d65fa5c5-8b38-4691-997b-7d159e6ba07d · outbound

This paper cites Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware Direct Preference Optimization.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware Direct Preference Optimization

Reference 99

Resolution: unresolved; no resolver link, observed 2026-08-10T14:29:28.337407Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 04c7a2c8-8f67-49cd-a605-bbb6ce9a015c · outbound

This paper cites Pre-trained multi- modal large language model enhances dermatological diagnosis using skingpt-4.

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink Pre-trained multi- modal large language model enhances dermatological diagnosis using skingpt-4

Reference 100

Resolution: verified fuzzy; raw_fallback, observed 2026-08-10T14:29:29.261895Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Pith citing papers

Observation d6d1e12e-8049-433f-a92f-0c05434cee76 · inbound

Extracting Visual Facts from Intermediate Layers for Mitigating Hallucinations in Multimodal Large Language Models cites this paper.

Extracting Visual Facts from Intermediate Layers for Mitigating Hallucinations in Multimodal Large Language Models Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-06T15:32:34.540866Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d790f951-692b-4272-a4bc-f4da2ba9d10c · inbound

ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments cites this paper.

ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 47

Resolution: verified exact; arxiv_id, observed 2026-05-19T01:02:54.889649Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 2ae76e75-cffa-4cab-9f63-b636d15419fb · inbound

Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement cites this paper.

Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-05-11T17:51:10.002380Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 6c4e377e-89a3-49e7-8b11-0c6de3a6bc17 · inbound

Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement cites this paper.

Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-05-12T07:06:35.968383Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 165690ed-78af-4d1a-add3-62a85499a74b · inbound

Babel: Jailbreaking Safety Attention via Obfuscation Distribution Optimized Sampling cites this paper.

Babel: Jailbreaking Safety Attention via Obfuscation Distribution Optimized Sampling Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-05-20T10:08:11.978912Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation a1e2cfed-d8cc-4f77-8f07-435cc7556f5a · inbound

Addressing Exacerbated Attention Sink for Source-Free Cross-Domain Few-Shot Learning cites this paper.

Addressing Exacerbated Attention Sink for Source-Free Cross-Domain Few-Shot Learning Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink

Reference 26

Resolution: verified exact; arxiv_id, observed 2026-06-29T22:34:01.808738Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.