Paper Citation Record · LEDGER

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection

As of 8 August 2026, this Paper Citation Record lists 20 of 20 outbound references and 0 inbound Pith citation observations for arXiv:2509.13608.

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

pith.paper-citation-record.v1
2509.13608 v2

Coverage vector

measured 20 of 20 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-04T16:30:29.434614Z

measured 20 of 20 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-08T06:32:00.761636+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

20 of 20 outbound references displayed

verified exact0
verified fuzzy0
unresolved20
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 1000e86f-98a5-4714-9d12-ac504861e9dc · outbound

This paper cites GPT-4o mini: advancing cost-efficient intelligence.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection GPT-4o mini: advancing cost-efficient intelligence

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.327387Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 74634001-9b30-416c-b525-439e772a006a · outbound

This paper cites ChatGPT Statistics 2025.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection ChatGPT Statistics 2025

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.378517Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 39719fdb-deff-4cba-b0b9-78b83161c184 · outbound

This paper cites The hateful memes challenge: Detecting hate speech in mul- timodal memes.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection The hateful memes challenge: Detecting hate speech in mul- timodal memes

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.427662Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3e9bfc75-46d5-40c5-b377-9e8566541b5f · outbound

This paper cites Cooperative in- verse reinforcement learning.Advances in neural information processing systems, 29, 2016.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Cooperative in- verse reinforcement learning.Advances in neural information processing systems, 29, 2016

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.511564Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 89eeb79d-e1ac-45fe-a168-562b51cb1e58 · outbound

This paper cites Viking, 2019.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Viking, 2019

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.570854Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6250ec90-87ae-4f03-91a9-c28a2c735827 · outbound

This paper cites BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.629504Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 59801848-7a23-4750-840d-8d662980be26 · outbound

This paper cites Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.691674Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e09ab6d1-fbea-4974-bb98-1d679ce53eff · outbound

This paper cites Singh, and V.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Singh, and V

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.740712Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d027d822-aaae-4b4f-90a2-0caecebd5e1c · outbound

This paper cites Multimodal detection of hateful memes by applying a vision-language pre-training model.PLoS one, 16(9):e0257302, 2021.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Multimodal detection of hateful memes by applying a vision-language pre-training model.PLoS one, 16(9):e0257302, 2021

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.803755Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6b1ac153-4b87-4865-9dfc-307f2ba5b179 · outbound

This paper cites Multimodal hate speech detection based on multi-task learning.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Multimodal hate speech detection based on multi-task learning

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.835192Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4767ef8c-c91b-4531-a401-6adbfcbdb9b6 · outbound

This paper cites Demystifying Hateful Content: Leveraging Large Multimodal Models for Hateful Meme Detection with Explainable Decisions.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Demystifying Hateful Content: Leveraging Large Multimodal Models for Hateful Meme Detection with Explainable Decisions

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.893257Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 72a80e90-7474-487a-9275-91f58e0d4e25 · outbound

This paper cites Toward Fairness via Maximum Mean Discrepancy Regularization on Logits Space.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Toward Fairness via Maximum Mean Discrepancy Regularization on Logits Space

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:28.962552Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2d31c79a-d3c7-462f-964f-b89a06b4b6b0 · outbound

This paper cites Classification of Carotid Plaque with Jellyfish Sign Through Convolutional and Recurrent Neural Networks Utilizing Plaque Surface Edges.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Classification of Carotid Plaque with Jellyfish Sign Through Convolutional and Recurrent Neural Networks Utilizing Plaque Surface Edges

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.002464Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 569bbb60-5c61-4576-ae5d-01e637352e70 · outbound

This paper cites AI ”safety” vs ”control” vs ”alignment”.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection AI ”safety” vs ”control” vs ”alignment”

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.084744Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7193b448-e006-4a89-8b57-be90ae8a648e · outbound

This paper cites How AI is shaping the next generation of brand safety and suitabil- ity.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection How AI is shaping the next generation of brand safety and suitabil- ity

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.134117Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4e3f0002-9a6e-4dd9-bfef-8b569985bc22 · outbound

This paper cites YesBut: A High-Quality Annotated Multimodal Dataset for evaluating Satire Comprehension capability of Vision-Language Models.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection YesBut: A High-Quality Annotated Multimodal Dataset for evaluating Satire Comprehension capability of Vision-Language Models

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.203246Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7b2d6dea-a34c-4679-af48-f1a3f2b8f599 · outbound

This paper cites SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Example.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Example

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.267971Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b8ba8047-e7dc-4314-9691-3abd4de223af · outbound

This paper cites Topologically Faithful Multi-class Segmentation in Medical Images.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Topologically Faithful Multi-class Segmentation in Medical Images

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.331818Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3894a040-3054-4e18-a724-87ac4b6121f9 · outbound

This paper cites Safety of Multimodal Large Language Models on Images and Texts.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Safety of Multimodal Large Language Models on Images and Texts

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.384240Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 305ba678-17fb-4ee7-891c-e7cc8f01e5e3 · outbound

This paper cites Safety of Multimodal Large Language Models on Images and Text.

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Safety of Multimodal Large Language Models on Images and Text

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-04T16:30:29.434614Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.