Paper Citation Record · LEDGER

See What You Are Told: Visual Attention Sink in Large Multimodal Models

As of 22 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 42 inbound Pith citation observations for arXiv:2503.03321.

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

pith.paper-citation-record.v1
2503.03321 v1

Coverage vector

measured 0 of 0 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links

measured 42 of 42 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-22T06:32:14.747728+00:00

measured 42 of 42 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-15T18:37:58.965659Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-07-04T03:19:31.045973Z

Reference resolution

0 of 0 outbound references displayed

verified exact0
verified fuzzy0
unresolved0
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

No outbound reference observations are available for this paper version.

Pith citing papers

Observation 163df10a-8836-4206-ae1b-7df13092ee82 · inbound

Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models cites this paper.

Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-09T15:26:18.833482Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 26f19368-69be-4297-ad00-dc9f67a84f3a · inbound

Seeing It or Not? Interpretable Vision-aware Latent Steering to Mitigate Object Hallucinations cites this paper.

Seeing It or Not? Interpretable Vision-aware Latent Steering to Mitigate Object Hallucinations See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-07T14:45:28.622401Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6765169-2bf6-441f-932a-d0bc6ad9374f · inbound

Not All Tokens and Heads Are Equally Important: Dual-Level Attention Intervention for Hallucination Mitigation cites this paper.

Not All Tokens and Heads Are Equally Important: Dual-Level Attention Intervention for Hallucination Mitigation See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-07T00:53:23.816759Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e4086921-1ee2-4750-a572-0c1d470f44b9 · inbound

PEVLM: Parallel Encoding for Vision-Language Models cites this paper.

PEVLM: Parallel Encoding for Vision-Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-15T18:37:58.965659Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 65ddff21-ce30-45c1-9483-4fd7cbd0887b · inbound

Examining Vision Language Models through Multi-dimensional Experiments with Vision and Text Features cites this paper.

Examining Vision Language Models through Multi-dimensional Experiments with Vision and Text Features See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-04T20:58:25.290792Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation faa9854e-6bc5-4828-aed1-bb21b25fe44d · inbound

HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling cites this paper.

HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 4

Resolution: verified exact; arxiv_id, observed 2026-05-21T21:14:22.812861Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 6e7dfff8-e5c2-4649-a669-34d764219291 · inbound

HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling cites this paper.

HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-04T14:42:57.009415Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3102a7f9-dd18-4a0b-b788-6034d468655c · inbound

Attention Misses Visual Risk: Risk-Adaptive Steering for Multimodal Safety Alignment cites this paper.

Attention Misses Visual Risk: Risk-Adaptive Steering for Multimodal Safety Alignment See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 2019

Resolution: unresolved; no resolver link, observed 2026-08-04T09:47:23.806393Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 863d9f65-ccbb-4e1e-9ae2-fd0b085c9787 · inbound

Capturing Gaze Shifts for Guidance: Cross-Modal Fusion Enhancement for VLM Hallucination Mitigation cites this paper.

Capturing Gaze Shifts for Guidance: Cross-Modal Fusion Enhancement for VLM Hallucination Mitigation See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-04T08:14:36.218482Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 465b55ba-ba7c-4563-a593-04e432d62705 · inbound

MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs cites this paper.

MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 27

Resolution: verified exact; arxiv_id, observed 2026-05-21T19:54:20.254402Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d1d25e1a-0373-4e3b-a027-a3115dc80d91 · inbound

MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs cites this paper.

MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-03T21:44:00.054831Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 36680381-e702-4798-bef3-21fc87835478 · inbound

Can Vision-Language Models Count? A Synthetic Benchmark and Analysis of Attention-Based Interventions cites this paper.

Can Vision-Language Models Count? A Synthetic Benchmark and Analysis of Attention-Based Interventions See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-17T20:10:11.265815Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 837504ba-23fd-4c0a-827d-e9080d9327e1 · inbound

EAGLE: Expert-Augmented Attention Guidance for Tuning-Free Industrial Anomaly Detection in Multimodal Large Language Models cites this paper.

EAGLE: Expert-Augmented Attention Guidance for Tuning-Free Industrial Anomaly Detection in Multimodal Large Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 18

Resolution: verified exact; arxiv_id, observed 2026-05-15T21:06:38.105307Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 019e877f-d788-46fd-a323-8e61986d8f3c · inbound

Deeper Thought, Weaker Aim: Understanding and Mitigating Perceptual Impairment during Reasoning in Multimodal Large Language Models cites this paper.

Deeper Thought, Weaker Aim: Understanding and Mitigating Perceptual Impairment during Reasoning in Multimodal Large Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-05-21T12:00:04.294537Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 34e3fb74-d1dc-4b4e-9f29-d0e0bfefbb2d · inbound

Counting to Four is still a Chore for VLMs cites this paper.

Counting to Four is still a Chore for VLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 17

Resolution: verified exact; arxiv_id, observed 2026-05-11T10:06:03.269739Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation f38698a3-d098-4a92-b8b8-cbf920c25201 · inbound

The Cost of Language: Centroid Erasure Exposes and Exploits Modal Competition in Multimodal Language Models cites this paper.

The Cost of Language: Centroid Erasure Exposes and Exploits Modal Competition in Multimodal Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-05-10T13:35:26.794760Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 2c89199c-f754-48a0-8a7c-b8b831309cbb · inbound

Sink-Token-Aware Pruning for Fine-Grained Video Understanding in Efficient Video LLMs cites this paper.

Sink-Token-Aware Pruning for Fine-Grained Video Understanding in Efficient Video LLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 17

Resolution: metadata mismatch; arxiv_id, observed 2026-05-10T00:44:48.397647Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 96e39971-fcc6-44f3-9e74-9d7d8f9df146 · inbound

Latent Denoising Improves Visual Alignment in Large Multimodal Models cites this paper.

Latent Denoising Improves Visual Alignment in Large Multimodal Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 40

Resolution: verified exact; arxiv_id, observed 2026-05-09T23:09:26.685247Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 15b448f1-d203-4a4e-871d-8f668ec1eb7d · inbound

Combating Visual Neglect and Semantic Drift in Large Multimodal Models for Enhanced Cross-Modal Retrieval cites this paper.

Combating Visual Neglect and Semantic Drift in Large Multimodal Models for Enhanced Cross-Modal Retrieval See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 17

Resolution: verified exact; arxiv_id, observed 2026-05-11T23:31:13.605771Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4566a0c8-e827-4395-bf5d-cb8ec66cd7a8 · inbound

Large Vision-Language Models Get Lost in Attention cites this paper.

Large Vision-Language Models Get Lost in Attention See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 27

Resolution: metadata mismatch; arxiv_id, observed 2026-05-11T19:26:10.066217Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation a3d5c165-df40-40d4-9ac9-b47f11a0ca60 · inbound

RAVE: Re-Allocating Visual Attention in Large Multimodal Models cites this paper.

RAVE: Re-Allocating Visual Attention in Large Multimodal Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-05-20T11:13:13.468869Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8f45ea5d-edad-4a0f-9064-b4ea98aaa6fe · inbound

RAVE: Re-Allocating Visual Attention in Large Multimodal Models cites this paper.

RAVE: Re-Allocating Visual Attention in Large Multimodal Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-07-01T14:55:48.355460Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 145b7c44-0cda-4d95-90c0-ff50baaa55d8 · inbound

MLLMs Know When Before Speaking: Revealing and Recovering Temporal Grounding via Attention Cues cites this paper.

MLLMs Know When Before Speaking: Revealing and Recovering Temporal Grounding via Attention Cues See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-22T07:14:42.456705Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 941f5569-b466-421a-8c5b-ffb153b9be89 · inbound

Inference Time Optimization with Confidence Dynamics cites this paper.

Inference Time Optimization with Confidence Dynamics See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-06-30T11:14:37.619616Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation f7cf9279-2f32-40c9-be26-fe63dbc86178 · 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 See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-06-29T22:34:01.800634Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation a41cd955-68d3-41d9-a696-11c4074591a9 · inbound

When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models cites this paper.

When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-07-02T01:56:27.409378Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b37d17ce-7aef-4aaa-8d64-58ba4f066f4e · inbound

Mechanistic Insights into Functional Sparsity in Multimodal LLMs via CoRe Heads cites this paper.

Mechanistic Insights into Functional Sparsity in Multimodal LLMs via CoRe Heads See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:46:56.644022Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b039fe84-d7ec-463f-a462-d74ffb2c4031 · inbound

Reason Twice: Segmentation via Candidate Discovery and Comparative Reasoning cites this paper.

Reason Twice: Segmentation via Candidate Discovery and Comparative Reasoning See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 26

Resolution: metadata mismatch; arxiv_id, observed 2026-07-03T00:47:30.634436Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation afb174be-9df7-440c-af56-d881b879226d · inbound

From Senses to Decisions: The Information Flow of Auditory and Visual Perception in Multimodal LLMs cites this paper.

From Senses to Decisions: The Information Flow of Auditory and Visual Perception in Multimodal LLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 20

Resolution: verified exact; arxiv_id, observed 2026-07-03T01:57:32.316937Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation f663055d-df4d-4e1e-b486-d9cf7ee7eb4a · inbound

Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models cites this paper.

Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 30

Resolution: verified exact; arxiv_id, observed 2026-07-03T11:28:04.108386Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d90ea871-600b-40cf-900d-1c1f81e05318 · inbound

Last But Not Least: Boundary Attention CalibratiON for Multimodal KV Cache Compression cites this paper.

Last But Not Least: Boundary Attention CalibratiON for Multimodal KV Cache Compression See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 38

Resolution: metadata mismatch; arxiv_id, observed 2026-07-03T09:17:49.042253Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation dbfcec07-4c4c-43c2-92b9-4538d13d9812 · inbound

The Hidden Evolution of Disguised Visual Context inside the VLM cites this paper.

The Hidden Evolution of Disguised Visual Context inside the VLM See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 25

Resolution: metadata mismatch; arxiv_id, observed 2026-07-04T03:19:31.048584Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 5082a377-8813-4fe3-ac05-7b19dd3bb3f9 · inbound

VisReflect: Latent Visual Reflection for Fine-Grained Perception in Long Visual Context cites this paper.

VisReflect: Latent Visual Reflection for Fine-Grained Perception in Long Visual Context See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 14

Resolution: metadata mismatch; arxiv_id, observed 2026-06-30T06:04:21.286659Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 9d60cd50-a443-4612-955b-80f85d85c4cc · inbound

ADAPT: Attention Dynamics Alignment with Preference Tuning for Faithful MLLMs cites this paper.

ADAPT: Attention Dynamics Alignment with Preference Tuning for Faithful MLLMs See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 10

Resolution: metadata mismatch; arxiv_id, observed 2026-07-01T06:55:29.486668Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 669815dd-0801-42ec-8b36-de554e29f9db · inbound

Information-Regularized Attention for Visual-Centric Reasoning cites this paper.

Information-Regularized Attention for Visual-Centric Reasoning See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-07-02T15:17:07.247821Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 6ff1c143-7362-4272-bccc-416129289b67 · inbound

The Ebb and Flow of Multimodal Focus: Scheduling Visual Relay Windows for Grounded VLM Reasoning cites this paper.

The Ebb and Flow of Multimodal Focus: Scheduling Visual Relay Windows for Grounded VLM Reasoning See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 18

Resolution: unresolved; no resolver link, observed 2026-07-14T05:37:19.632869Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e2cba484-7271-45db-ada6-1c7465442908 · inbound

The Ebb and Flow of Multimodal Focus: Scheduling Visual Relay Windows for Grounded VLM Reasoning cites this paper.

The Ebb and Flow of Multimodal Focus: Scheduling Visual Relay Windows for Grounded VLM Reasoning See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-02T07:01:19.935203Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dfaf8944-d82a-4aba-bcd1-997dde3b7c7d · inbound

DrawingVQA: A Real-World Benchmark for Multi-Depth Visual-Textual Reasoning on Construction Drawings cites this paper.

DrawingVQA: A Real-World Benchmark for Multi-Depth Visual-Textual Reasoning on Construction Drawings See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-01T23:29:00.730157Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a9d423ba-0d22-4dc5-b4a2-33fed6b76e9e · inbound

ST-Veto: Spatio-Temporal Token Veto for Diffusion MLLMs via Taylor Prediction and Visual Grounding cites this paper.

ST-Veto: Spatio-Temporal Token Veto for Diffusion MLLMs via Taylor Prediction and Visual Grounding See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-01T16:48:35.906973Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b5b66a19-8b7a-43b8-b13f-0909d0cc1a4a · inbound

Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers cites this paper.

Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-01T13:25:54.711000Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cd3f7e0f-08b8-4be6-bc56-4994d17e9b8b · inbound

Not All Redundant Tokens Are Alike: Analyzing Visual Token Pruning through Token Roles cites this paper.

Not All Redundant Tokens Are Alike: Analyzing Visual Token Pruning through Token Roles See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-06T23:28:37.236788Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc3af201-0a87-4966-a990-2f6c2f8c9e5d · inbound

Same Attention, Different Truths: Put Logit-Lens over Visual Attention to Detect and Mitigate LVLM Object Hallucination cites this paper.

Same Attention, Different Truths: Put Logit-Lens over Visual Attention to Detect and Mitigate LVLM Object Hallucination See What You Are Told: Visual Attention Sink in Large Multimodal Models

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-10T10:47:43.957166Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.