🔄 Transformers - Bingran · Scour

Transformer-Enhanced Reinforcement Learning: Fundamentals and Applications in Communication Networks

🎮Reinforcement Learning Academic

SpikeDecoder: Realizing the GPT Architecture with Spiking Neural Networks

🧠AI Research Academic

Chiaroscuro Attention: Spending Compute in the Dark

📉Deep Learning Academic

PT-WNO: Point Transformer with Wavelet Neural Operator for 3D Point Cloud Semantic Segmentation

📐Scaling Laws Academic

Look Less, Reason More: Block-wise Attention Skipping for Efficient Multimodal LLMs

💬LLMs Academic

Towards Tight Bounds for Streaming Attention

🧠AI Research Academic

UR-BERT: Scaling Text Encoders for Massively Multilingual TTS Through Universal Romanization and Speech Token Prediction

💬LLMs Academic

LiteVSR: Lightweight Adaptation of Frozen Diffusion Transformers for Video Super-Resolution

⚙️Model Training Academic

Kuramoto Attention: Synchronizing Self-Attention on the Torus

📐Scaling Laws Academic

ATT-CR: Adaptive Triangular Transformer for Cloud Removal

📐Scaling Laws Academic

TextEconomizer: Enhancing Lossy Text Compression with Denoising Transformers and Entropy Coding

💬LLMs Academic

SPADE: Split-and-Delay Embeddings for Autoregressive High-Granularity Calorimeter Simulation

🧠AI Research Academic

From Architecture to Output: Structural Origins of Hallucination in Large Language Models and the Amplifying Role of Data

💬LLMs Academic

LifeSentence: Language models can encode human life course trajectories from longitudinal panel data

🧠AI Research Academic

Gated Bidirectional Linear Attention for Generative Retrieval

💬LLMs Academic

RePAIR: Predictive Self-Supervised Representation Learning in Chess

🎮Reinforcement Learning Academic

End-to-End Context Compression at Scale

⚙️Model Training Academic

Architecture-Aware Reinforcement Learning Makes Sliding-Window Attention Competitive in Math Reasoning

⚙️Model Training Academic

GRAMformer: Any-Order Modality Interactions via Volumetric Multimodal Cross-Attention

🧠AI Research Academic

NGram-MoSE: Efficient Remote Sensing Super-Resolution via N-Gram Context and Mixture-of-Experts

🧠AI Research Academic

Log in to enable infinite scrolling