A regime theory selects the optimal controller class for LLM action decisions from a nested lattice of four classes using three data-estimable bottlenecks, with a Bernstein-tight threshold and empirical matches on multiple benchmarks.
hub
Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=
11 Pith papers cite this work. Polarity classification is still indexing.
11
Pith papers citing it
hub tools
representative citing papers
ReaLB balances multimodal MoE inference loads by switching vision-heavy experts to lower FP4 precision per device rank, hiding the change in the dispatch phase to deliver 1.10-1.32x speedup with <1% accuracy degradation.
MMGuard generates unlearnable multimodal examples via perturbations that exploit LVLM optimization shortcuts and disrupt cross-modal bindings, providing robust protection against unauthorized fine-tuning across threat models.
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.
In LVLMs, attention can be replaced by random Gaussian weights with little or no performance loss, indicating that current models get lost in attention rather than efficiently using visual context.
RetentiveKV uses entropy to drive state-space model transitions that retain and reactivate low-attention visual tokens in a continuous memory instead of pruning them, delivering 5x KV cache compression and 1.5x faster decoding.
DoRA improves LoRA by decomposing weights into magnitude and direction and updating only direction with low-rank matrices, closing much of the gap to full fine-tuning.
Video-LLaVA creates a unified visual representation for images and videos via pre-projection alignment, enabling mutual enhancement from joint training and strong results on image and video benchmarks.
mPLUG-Owl3 introduces hyper attention blocks to integrate vision and language for long image-sequence understanding and reports SOTA results on single-image, multi-image, and video benchmarks.
The paper defines Agent AI as interactive multimodal systems that perceive grounded data and generate embodied actions, arguing this approach can mitigate hallucinations in foundation models.
citing papers explorer
-
A Regime Theory of Controller Class Selection for LLM Action Decisions
A regime theory selects the optimal controller class for LLM action decisions from a nested lattice of four classes using three data-estimable bottlenecks, with a Bernstein-tight threshold and empirical matches on multiple benchmarks.
-
ReaLB: Real-Time Load Balancing for Multimodal MoE Inference
ReaLB balances multimodal MoE inference loads by switching vision-heavy experts to lower FP4 precision per device rank, hiding the change in the dispatch phase to deliver 1.10-1.32x speedup with <1% accuracy degradation.
-
To See is Not to Learn: Protecting Multimodal Data from Unauthorized Fine-Tuning of Large Vision-Language Model
MMGuard generates unlearnable multimodal examples via perturbations that exploit LVLM optimization shortcuts and disrupt cross-modal bindings, providing robust protection against unauthorized fine-tuning across threat models.
-
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.
-
Large Vision-Language Models Get Lost in Attention
In LVLMs, attention can be replaced by random Gaussian weights with little or no performance loss, indicating that current models get lost in attention rather than efficiently using visual context.
-
RetentiveKV: State-Space Memory for Uncertainty-Aware Multimodal KV Cache Eviction
RetentiveKV uses entropy to drive state-space model transitions that retain and reactivate low-attention visual tokens in a continuous memory instead of pruning them, delivering 5x KV cache compression and 1.5x faster decoding.
-
DoRA: Weight-Decomposed Low-Rank Adaptation
DoRA improves LoRA by decomposing weights into magnitude and direction and updating only direction with low-rank matrices, closing much of the gap to full fine-tuning.
-
Video-LLaVA: Learning United Visual Representation by Alignment Before Projection
Video-LLaVA creates a unified visual representation for images and videos via pre-projection alignment, enabling mutual enhancement from joint training and strong results on image and video benchmarks.
-
mPLUG-Owl3: Towards Long Image-Sequence Understanding in Multi-Modal Large Language Models
mPLUG-Owl3 introduces hyper attention blocks to integrate vision and language for long image-sequence understanding and reports SOTA results on single-image, multi-image, and video benchmarks.
-
Agent AI: Surveying the Horizons of Multimodal Interaction
The paper defines Agent AI as interactive multimodal systems that perceive grounded data and generate embodied actions, arguing this approach can mitigate hallucinations in foundation models.
- MACS: Modality-Aware Capacity Scaling for Efficient Multimodal MoE Inference