Massive activations first appear in a single ME Layer due to RMSNorm and FFN, remain invariant thereafter, and a simple softening method raises LLM performance while reducing attention sinks.
hub
arXiv preprint arXiv:2402.17762 (2024) 18 J
17 Pith papers cite this work. Polarity classification is still indexing.
17
Pith papers citing it
hub tools
representative citing papers
Sink-Token-aware Pruning (SToP) suppresses semantically uninformative sink tokens during visual token pruning in Video LLMs, boosting fine-grained performance even at 90% pruning rates across hallucination, reasoning, and MCQA benchmarks.
Attention sinks in LVLM create a global-vs-local trade-off that a layer-wise gating module can balance to improve multimodal benchmark performance.
K-sparse autoencoders with dead-latent fixes produce clean scaling laws and better feature quality metrics that improve with size, shown by training a 16-million-latent model on GPT-4 activations.
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.
Suppressing attention sinks in diffusion transformers does not degrade CLIP-T alignment at moderate levels but induces sink-specific perceptual shifts six times larger than equal-budget random masking.
Outlier tokens in DiTs are addressed with Dual-Stage Registers, which reduce artifacts and improve image generation on ImageNet and text-to-image tasks.
TIGS detects backdoor-induced attention collapse in LLMs and applies content-aware tail-risk screening plus intrinsic geometric smoothing to suppress attacks while preserving normal performance.
GRACE reframes KV cache channel pruning as graph optimization to find a near-optimal subset, achieving 60% compression with negligible degradation and outperforming prior methods.
Applying a head-specific sigmoid gate after SDPA in LLMs boosts performance and stability by adding non-linearity and query-dependent sparse modulation while reducing attention sinks.
PyramidKV dynamically compresses KV cache across layers following pyramidal information funneling, matching full performance at 12% retention and outperforming alternatives at 0.7% retention with up to 20.5 accuracy gains.
HyperLens reveals that deeper transformer layers magnify small confidence changes into fine-grained trajectories, allowing quantification of cognitive effort where complex tasks demand more and standard SFT can reduce it.
Colinearity-Decay regularizer trains ViTs that maintain or improve full-precision accuracy while delivering higher accuracy after low-bit quantization on ImageNet and COCO tasks.
OSC separates token-persistent outlier channels in activations into a compact high-precision tensor for dual-path 4-bit GEMM computation, limiting accuracy loss to roughly 1-2 points on Qwen3 models while delivering up to 1.78x speedup over W8A8 baselines.
Residual-stream noise injection raises narrative diversity in Arabic educational stories while preserving reading-grade level, outperforming high-temperature sampling across five 7-9B models.
MiMo-V2-Flash is a 309B/15B MoE model trained on 27T tokens with hybrid attention and multi-teacher on-policy distillation that matches larger models like DeepSeek-V3.2 while enabling 2.6x faster decoding via repurposed MTP layers.
DuQuant++ adapts outlier-aware fine-grained rotation to MXFP4 by matching block size to the 32-element microscaling group, enabling a single rotation that smooths distributions and achieves SOTA performance on LLaMA-3 with lower cost.
citing papers explorer
-
A Single Layer to Explain Them All:Understanding Massive Activations in Large Language Models
Massive activations first appear in a single ME Layer due to RMSNorm and FFN, remain invariant thereafter, and a simple softening method raises LLM performance while reducing attention sinks.
-
Sink-Token-Aware Pruning for Fine-Grained Video Understanding in Efficient Video LLMs
Sink-Token-aware Pruning (SToP) suppresses semantically uninformative sink tokens during visual token pruning in Video LLMs, boosting fine-grained performance even at 90% pruning rates across hallucination, reasoning, and MCQA benchmarks.
-
When Sinks Help or Hurt: Unified Framework for Attention Sink in Large Vision-Language Models
Attention sinks in LVLM create a global-vs-local trade-off that a layer-wise gating module can balance to improve multimodal benchmark performance.
-
Scaling and evaluating sparse autoencoders
K-sparse autoencoders with dead-latent fixes produce clean scaling laws and better feature quality metrics that improve with size, shown by training a 16-million-latent model on GPT-4 activations.
-
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.
-
Attention Sinks in Diffusion Transformers: A Causal Analysis
Suppressing attention sinks in diffusion transformers does not degrade CLIP-T alignment at moderate levels but induces sink-specific perceptual shifts six times larger than equal-budget random masking.
-
Taming Outlier Tokens in Diffusion Transformers
Outlier tokens in DiTs are addressed with Dual-Stage Registers, which reduce artifacts and improve image generation on ImageNet and text-to-image tasks.
-
Defusing the Trigger: Plug-and-Play Defense for Backdoored LLMs via Tail-Risk Intrinsic Geometric Smoothing
TIGS detects backdoor-induced attention collapse in LLMs and applies content-aware tail-risk screening plus intrinsic geometric smoothing to suppress attacks while preserving normal performance.
-
Graph-Guided Adaptive Channel Elimination for KV Cache Compression
GRACE reframes KV cache channel pruning as graph optimization to find a near-optimal subset, achieving 60% compression with negligible degradation and outperforming prior methods.
-
Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink-Free
Applying a head-specific sigmoid gate after SDPA in LLMs boosts performance and stability by adding non-linearity and query-dependent sparse modulation while reducing attention sinks.
-
PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
PyramidKV dynamically compresses KV cache across layers following pyramidal information funneling, matching full performance at 12% retention and outperforming alternatives at 0.7% retention with up to 20.5 accuracy gains.
-
HyperLens: Quantifying Cognitive Effort in LLMs with Fine-grained Confidence Trajectory
HyperLens reveals that deeper transformer layers magnify small confidence changes into fine-grained trajectories, allowing quantification of cognitive effort where complex tasks demand more and standard SFT can reduce it.
-
Colinearity Decay: Training Quantization-Friendly ViTs with Outlier Decay
Colinearity-Decay regularizer trains ViTs that maintain or improve full-precision accuracy while delivering higher accuracy after low-bit quantization on ImageNet and COCO tasks.
-
OSC: Hardware Efficient W4A4 Quantization via Outlier Separation in Channel Dimension
OSC separates token-persistent outlier channels in activations into a compact high-precision tensor for dual-path 4-bit GEMM computation, limiting accuracy loss to roughly 1-2 points on Qwen3 models while delivering up to 1.78x speedup over W8A8 baselines.
-
Noise Steering for Controlled Text Generation: Improving Diversity and Reading-Level Fidelity in Arabic Educational Story Generation
Residual-stream noise injection raises narrative diversity in Arabic educational stories while preserving reading-grade level, outperforming high-temperature sampling across five 7-9B models.
-
MiMo-V2-Flash Technical Report
MiMo-V2-Flash is a 309B/15B MoE model trained on 27T tokens with hybrid attention and multi-teacher on-policy distillation that matches larger models like DeepSeek-V3.2 while enabling 2.6x faster decoding via repurposed MTP layers.
-
DuQuant++: Fine-grained Rotation Enhances Microscaling FP4 Quantization
DuQuant++ adapts outlier-aware fine-grained rotation to MXFP4 by matching block size to the 32-element microscaling group, enabling a single rotation that smooths distributions and achieves SOTA performance on LLaMA-3 with lower cost.