Lior Wolf

Factor Graph Optimization for Belief Propagation Decoding

On the Recall Scaling Laws in Mamba: A Theoretical and Mechanistic Study via Hashing

VideoMB: Steering Representations towards Motion Balanced Caption Generation in Vision-Language Models

Deep Active Speech Cancellation with Mamba-Masking Network

Overclocking LLM Reasoning: Monitoring and Controlling Thinking Path Lengths in LLMs

Query-Guided Prototype Generation for Few-Class Classification

Execution Guided Line-by-Line Code Generation

VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models

FlowMo: Variance-Based Flow Guidance for Coherent Motion in Video Generation

Detecting and Pruning Prominent but Detrimental Neurons in Large Language Models

Explaining Modern Gated-Linear RNNs via a Unified Implicit Attention Formulation

Overflow Prevention Enhances Long-Context Recurrent LLMs

Accelerating Error Correction Code Transformers

Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Models

Revisiting LRP: Positional Attribution as the Missing Ingredient for Transformer Explainability

Active Audio Cancellation with Multi-Band Mamba Network

Mitigating Copy Bias in In-Context Learning through Neuron Pruning

DeciMamba: Exploring the Length Extrapolation Potential of Mamba

Factor Graph Optimization of Error-Correcting Codes for Belief Propagation Decoding