·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Neil Rimer thinks the AI money is coming back out2h◆The Steering Budget: Examples beat Knobs3h◆Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs3h◆RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination3h◆When a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models3h◆SAGA: Schema-Aware Grounding for Agentic Text-to-SPARQL Generation3h◆ReasFlow: Assisting Reasoning-Centric Scientific Discovery in Applied Mathematics via a Knowledge-Based Multi-Agent System3h◆LBA: Textual Hard-Label Adversarial Attack under Low Query Budgets3h◆Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utility3h◆Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation3h◆LIGO-PINN: Learned Initialization via Gated Optimization to Alleviate Convergence Failures in Physics Informed Neural Networks3h◆OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios3h◆When AI Blurs the Boundaries of Contribution: An Empirical Study of Authorship Calibration3h◆Digital Pantheon: Simulating and Auditing Coalition Formation with LLM Agents3h◆Inoculation Adapters: Improved Selective Generalization of Capabilities with Fewer Surprising Backdoors3h◆ANet Patu-1: The Value of Connection in the Agent Network3h◆Automatically Evolving Prompt Guidelines for Task-Specific Optimization3h◆MAPS: Modeling Co-Existing Subjective Perspectives and Shared Meaning in Multi-Agent Cognitive Dialogue3h◆Automatic Hard Example Synthesis with Multi-Level Agentic Data Curation3h◆Tracing LLM Behavior to the Training Data with Empirical Next-Token Distributions3h◆Neil Rimer thinks the AI money is coming back out2h◆The Steering Budget: Examples beat Knobs3h◆Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs3h◆RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination3h◆When a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models3h◆SAGA: Schema-Aware Grounding for Agentic Text-to-SPARQL Generation3h◆ReasFlow: Assisting Reasoning-Centric Scientific Discovery in Applied Mathematics via a Knowledge-Based Multi-Agent System3h◆LBA: Textual Hard-Label Adversarial Attack under Low Query Budgets3h◆Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utility3h◆Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation3h◆LIGO-PINN: Learned Initialization via Gated Optimization to Alleviate Convergence Failures in Physics Informed Neural Networks3h◆OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios3h◆When AI Blurs the Boundaries of Contribution: An Empirical Study of Authorship Calibration3h◆Digital Pantheon: Simulating and Auditing Coalition Formation with LLM Agents3h◆Inoculation Adapters: Improved Selective Generalization of Capabilities with Fewer Surprising Backdoors3h◆ANet Patu-1: The Value of Connection in the Agent Network3h◆Automatically Evolving Prompt Guidelines for Task-Specific Optimization3h◆MAPS: Modeling Co-Existing Subjective Perspectives and Shared Meaning in Multi-Agent Cognitive Dialogue3h◆Automatic Hard Example Synthesis with Multi-Level Agentic Data Curation3h◆Tracing LLM Behavior to the Training Data with Empirical Next-Token Distributions3h◆
News/Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection
arxiv
PublishedMay 26, 2026 at 4:00 AM

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

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2509.13608v2 Announce Type: replace Abstract: As Large Multimodal Models (LMMs) become integral to daily digital life, understanding their safety architectures is a critical problem for AI Alignment. This paper presents a systematic analysis of OpenAI's GPT-4o mini, a globally deployed model,

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivThe Steering Budget: Examples beat Knobs3harxivPolestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs3harxivRxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination3harxivWhen a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models3h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews