LTX-2: Efficient Joint Audio-Visual Foundation Model
Paper • 2601.03233 • Published • 176
Stuff I'm going to read
-
-
MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head
Paper • 2601.07832 • Published • 52 -
Motion Attribution for Video Generation
Paper • 2601.08828 • Published • 71 -
Post-LayerNorm Is Back: Stable, ExpressivE, and Deep
Paper • 2601.19895 • Published • 26 -
Elastic Attention: Test-time Adaptive Sparsity Ratios for Efficient Transformers
Paper • 2601.17367 • Published • 34 -
Advancing Open-source World Models
Paper • 2601.20540 • Published • 134 -
Why Attention Patterns Exist: A Unifying Temporal Perspective Analysis
Paper • 2601.21709 • Published • 3 -
ERNIE 5.0 Technical Report
Paper • 2602.04705 • Published • 266 -
FASA: Frequency-aware Sparse Attention
Paper • 2602.03152 • Published • 153 -
LLaDA2.1: Speeding Up Text Diffusion via Token Editing
Paper • 2602.08676 • Published • 70 -
MOVA: Towards Scalable and Synchronized Video-Audio Generation
Paper • 2602.08794 • Published • 158 -
OPUS: Towards Efficient and Principled Data Selection in Large Language Model Pre-training in Every Iteration
Paper • 2602.05400 • Published • 350 -
When to Memorize and When to Stop: Gated Recurrent Memory for Long-Context Reasoning
Paper • 2602.10560 • Published • 31 -
Towards Autonomous Mathematics Research
Paper • 2602.10177 • Published • 36 -
MOSS-Audio-Tokenizer: Scaling Audio Tokenizers for Future Audio Foundation Models
Paper • 2602.10934 • Published • 49 -
Experiential Reinforcement Learning
Paper • 2602.13949 • Published • 72 -
BitDance: Scaling Autoregressive Generative Models with Binary Tokens
Paper • 2602.14041 • Published • 53 -
STAPO: Stabilizing Reinforcement Learning for LLMs by Silencing Rare Spurious Tokens
Paper • 2602.15620 • Published • 3 -
SLA2: Sparse-Linear Attention with Learnable Routing and QAT
Paper • 2602.12675 • Published • 58 -
VESPO: Variational Sequence-Level Soft Policy Optimization for Stable Off-Policy LLM Training
Paper • 2602.10693 • Published • 220 -
Avey-B
Paper • 2602.15814 • Published • 3 -
Decoding as Optimisation on the Probability Simplex: From Top-K to Top-P (Nucleus) to Best-of-K Samplers
Paper • 2602.18292 • Published • 12 -
Test-Time Training with KV Binding Is Secretly Linear Attention
Paper • 2602.21204 • Published • 31 -
Memory Caching: RNNs with Growing Memory
Paper • 2602.24281 • Published • 9 -
Timer-S1: A Billion-Scale Time Series Foundation Model with Serial Scaling
Paper • 2603.04791 • Published • 17 -
LoGeR: Long-Context Geometric Reconstruction with Hybrid Memory
Paper • 2603.03269 • Published • 62 -
ReMix: Reinforcement routing for mixtures of LoRAs in LLM finetuning
Paper • 2603.10160 • Published • 25 -
V_{0.5}: Generalist Value Model as a Prior for Sparse RL Rollouts
Paper • 2603.10848 • Published • 14 -
Video-Based Reward Modeling for Computer-Use Agents
Paper • 2603.10178 • Published • 43 -
Attention Residuals
Paper • 2603.15031 • Published • 176 -
Efficient Exploration at Scale
Paper • 2603.17378 • Published • 13 -
MinerU-Diffusion: Rethinking Document OCR as Inverse Rendering via Diffusion Decoding
Paper • 2603.22458 • Published • 134 -
Why Does Self-Distillation (Sometimes) Degrade the Reasoning Capability of LLMs?
Paper • 2603.24472 • Published • 51 -
Voxtral TTS
Paper • 2603.25551 • Published • 57 -
Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models
Paper • 2603.25716 • Published • 152 -
LongCat-Next: Lexicalizing Modalities as Discrete Tokens
Paper • 2603.27538 • Published • 137 -
TriAttention: Efficient Long Reasoning with Trigonometric KV Compression
Paper • 2604.04921 • Published • 83