20 citations · 41 across the 52 of their papers we have counts for
58 papers
Depth Memory
David González-Martínez, Philipp Nazari, T. Konstantin Rusch +1
Residual connections propagate information through a network by additively accumulating layer outputs. Recently proposed alternatives, such as Manifold-Constrained Hyper-Connection…
ISO: An RLVR-Native Optimization Stack
Hanqing Zhu, Wenyan Cong, Zhizhou Sha +8
Reinforcement learning with verifiable rewards (RLVR) advances the reasoning capabilities of language models, yet the optimization layer that converts reward feedback into weight-s…
AlphaQ: Calibration-Free Bit Allocation for Mixture-of-Experts Quantization
Wanqi Yang, Yuexiao Ma, Alexander Conzelmann +4
Mixture-of-Experts (MoE) architectures scale model capacity through sparse expert activation, but their deployment remains memory-bound because all expert weights must reside in me…
One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs
Di He, Songjun Tu, Keyu Wang +2
Learning rate configuration is a fundamental aspect of modern deep learning. The prevailing practice of applying a uniform learning rate across all layers overlooks the structural…
Motion-Aware Caching for Efficient Autoregressive Video Generation
Jing Xu, Yuexiao Ma, Xuzhe Zheng +7
Autoregressive video generation paradigms offer theoretical promise for long video synthesis, yet their practical deployment is hindered by the computational burden of sequential i…
Layer Collapse in Diffusion Language Models
Alexander Conzelmann, Albert Catalan-Tatjer, Shiwei Liu
Diffusion language models (DLMs) have recently emerged as competitive alternatives to autoregressive (AR) language models, yet differences in their activation dynamics remain poorl…
ELAS: Efficient Pre-Training of Low-Rank Large Language Models via 2:4 Activation Sparsity
Jiaxi Li, Lu Yin, Li Shen +5
Large Language Models (LLMs) have achieved remarkable capabilities, but their immense computational demands during training remain a critical bottleneck for widespread adoption. Lo…
When Does Sparsity Mitigate the Curse of Depth in LLMs
Dilxat Muhtar, Xinyuan Song, Sebastian Pokutta +4
Recent work has demonstrated the curse of depth in large language models (LLMs), where later layers contribute less to learning and representation than earlier layers. Such under-u…
Why Diffusion Language Models Struggle with Truly Parallel (Non-Autoregressive) Decoding?
Pengxiang Li, Dilxat Muhtar, Tianlong Chen +2
Diffusion Language Models (DLMs) are often advertised as enabling parallel token generation, yet practical fast DLMs frequently converge to left-to-right, autoregressive (AR)-like…
Search or Accelerate: Confidence-Switched Position Beam Search for Diffusion Language Models
Mingyu Cao, Alvaro H. C. Correia, Christos Louizos +2
Diffusion Language Models (DLMs) generate text by iteratively denoising a masked sequence, repeatedly deciding which positions to commit at each step. Standard decoding follows a g…
ActTail: Global Activation Sparsity in Large Language Models
Wenwen Hou, Xinyuan Song, Shiwei Liu
Activation sparsity is a promising approach for accelerating large language model (LLM) inference by reducing computation and memory movement. However, existing activation sparsity…
Rewiring Experts on the Fly:Continuous Rerouting for Better Online Adaptation in Mixture-of-Expert models
Guinan Su, Yanwu Yang, Li Shen +3
Mixture-of-Experts (MoE) models achieve efficient scaling through sparse expert activation, but often suffer from suboptimal routing decisions due to distribution shifts in deploym…
Neural Sum-of-Squares: Certifying the Nonnegativity of Polynomials with Transformers
Nico Pelleriti, Christoph Spiegel, Shiwei Liu +3
Certifying nonnegativity of polynomials is a well-known NP-hard problem with direct applications spanning non-convex optimization, control, robotics, and beyond. A sufficient condi…
When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs
Keyu Wang, Tian Lyu, Guinan Su +4
Layer pruning has emerged as a widely adopted technique for improving the efficiency of large language models (LLMs). Although existing methods demonstrate strong performance reten…
Demystifying the Roles of LLM Layers in Retrieval, Knowledge, and Reasoning
Xinyuan Song, Keyu Wang, PengXiang Li +2
Recent studies suggest that the deeper layers of Large Language Models (LLMs) contribute little to representation learning and can often be removed without significant performance…
Model Unmerging: Making Your Models Unmergeable for Secure Model Sharing
Zihao Wang, Enneng Yang, Lu Yin +2
Model merging leverages multiple finetuned expert models to construct a multi-task model with low cost, and is gaining increasing attention. However, as a growing number of finetun…
LOST: Low-rank and Sparse Pre-training for Large Language Models
Jiaxi Li, Lu Yin, Li Shen +6
While large language models (LLMs) have achieved remarkable performance across a wide range of tasks, their massive scale incurs prohibitive computational and memory costs for pre-…
Diffusion Language Models Know the Answer Before Decoding
Pengxiang Li, Yefan Zhou, Dilxat Muhtar +5
Diffusion language models (DLMs) have recently emerged as an alternative to autoregressive approaches, offering parallel sequence generation and flexible token orders. However, the…
GPAS: Accelerating Convergence of LLM Pretraining via Gradient-Preserving Activation Scaling
Tianhao Chen, Xin Xu, Zijing Liu +12
Modern Large Language Models, such as the LLaMA, Qwen and DeepSeek series, predominantly adopt the Pre-LayerNorm (Pre-LN) Transformer architecture. While being stable during pretra…
GPTailor: Large Language Model Pruning Through Layer Cutting and Stitching
Guinan Su, Li Shen, Lu Yin +3
Large language models (LLMs) have shown remarkable capabilities in language understanding and generation. However, such impressive capability typically comes with a substantial mod…
Chain-of-Experts: Unlocking the Communication Power of Mixture-of-Experts Models
Zihan Wang, Rui Pan, Jiarui Yao +7
We propose Chain-of-Experts (CoE), a new Mixture-of-Experts (MoE) architecture that introduces sequential expert communication within each layer. Unlike traditional MoE models, whe…
Double-Checker: Enhancing Reasoning of Slow-Thinking LLMs via Self-Critical Fine-Tuning
Xin Xu, Tianhao Chen, Fan Zhang +11
While slow-thinking large language models (LLMs) exhibit reflection-like reasoning, commonly referred to as the "aha moment:, their ability to generate informative critiques and re…
AlphaDecay: Module-wise Weight Decay for Heavy-Tailed Balancing in LLMs
Di He, Songjun Tu, Ajay Jaiswal +4
Weight decay is a standard regularization technique for training large language models (LLMs). While it is common to assign a uniform decay rate to every layer, this approach overl…
Effective Learning for Small Reasoning Models: An Empirical Study on 0.5B Reasoning LLMs
Xialie Zhuang, Peixian Ma, Zhikai Jia +2
The ongoing evolution of language models has led to the development of large-scale architectures that demonstrate exceptional performance across a wide range of tasks. However, the…