KoALa-Bench is a new public benchmark with six tasks that tests Korean speech recognition, translation, question answering, instruction following, and faithfulness in large audio language models.
hub
Halle-switch: Rethinking and controlling ob- ject existence hallucinations in large vision language mod- els for detailed caption
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
hub tools
citation-role summary
method 1
citation-polarity summary
roles
method 1polarities
background 1representative citing papers
CaptionQA is a new benchmark with 33,027 questions across natural, document, e-commerce, and embodied AI domains that measures how much utility model-generated captions retain compared to original images when used by LLMs for downstream tasks.
Proposes CSR task and HalluSegBench using visual counterfactuals to diagnose segmentation hallucinations in VLMs, plus RobustSeg via counterfactual fine-tuning that reduces hallucinations by 30% on FP-RefCOCO.
ZINA detects fine-grained hallucinations in MLLM outputs, classifies errors into six types, and proposes edits, outperforming GPT-4o and Llama-3.2 on the new VisionHall dataset of annotated and synthetic samples.
Reweighting training emphasis toward image-negative tokens and filtering hallucinated data reduces object hallucination in LVLMs across three model variants.
Deep Pre-Alignment uses a small VLM perceiver instead of ViT to pre-align visual features with LLM text space, yielding 1.9-3.0 point gains on multimodal benchmarks and 32.9% less language forgetting.
LVLMs show vocabulary hijacking by inert tokens that decode to hijacking anchors; HABI locates them, NHAR finds resilient heads, and HAVAE boosts those heads to cut hallucinations.
AMBER is an LLM-free multi-dimensional benchmark for evaluating hallucinations in MLLMs across generative and discriminative tasks.
This survey reviews the definition, symptoms, evaluation benchmarks, root causes, and mitigation methods for hallucinations in large vision-language models.
This survey organizes the architectures, training strategies, data, evaluation methods, extensions, and challenges of Multimodal Large Language Models.
citing papers explorer
-
KoALa-Bench: Evaluating Large Audio Language Models on Korean Speech Understanding and Faithfulness
KoALa-Bench is a new public benchmark with six tasks that tests Korean speech recognition, translation, question answering, instruction following, and faithfulness in large audio language models.
-
CaptionQA: Is Your Caption as Useful as the Image Itself?
CaptionQA is a new benchmark with 33,027 questions across natural, document, e-commerce, and embodied AI domains that measures how much utility model-generated captions retain compared to original images when used by LLMs for downstream tasks.
-
Counterfactual Segmentation Reasoning: Diagnosing and Mitigating Pixel-Grounding Hallucination
Proposes CSR task and HalluSegBench using visual counterfactuals to diagnose segmentation hallucinations in VLMs, plus RobustSeg via counterfactual fine-tuning that reduces hallucinations by 30% on FP-RefCOCO.
-
ZINA: Multimodal Fine-grained Hallucination Detection and Editing
ZINA detects fine-grained hallucinations in MLLM outputs, classifies errors into six types, and proposes edits, outperforming GPT-4o and Llama-3.2 on the new VisionHall dataset of annotated and synthetic samples.
-
Reducing Object Hallucination in LVLMs via Emphasizing Image-negative Tokens
Reweighting training emphasis toward image-negative tokens and filtering hallucinated data reduces object hallucination in LVLMs across three model variants.
-
Deep Pre-Alignment for VLMs
Deep Pre-Alignment uses a small VLM perceiver instead of ViT to pre-align visual features with LLM text space, yielding 1.9-3.0 point gains on multimodal benchmarks and 32.9% less language forgetting.
-
Vocabulary Hijacking in LVLMs: Unveiling Critical Attention Heads by Excluding Inert Tokens to Mitigate Hallucination
LVLMs show vocabulary hijacking by inert tokens that decode to hijacking anchors; HABI locates them, NHAR finds resilient heads, and HAVAE boosts those heads to cut hallucinations.
-
AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation
AMBER is an LLM-free multi-dimensional benchmark for evaluating hallucinations in MLLMs across generative and discriminative tasks.
-
A Survey on Hallucination in Large Vision-Language Models
This survey reviews the definition, symptoms, evaluation benchmarks, root causes, and mitigation methods for hallucinations in large vision-language models.
-
A Survey on Multimodal Large Language Models
This survey organizes the architectures, training strategies, data, evaluation methods, extensions, and challenges of Multimodal Large Language Models.