{"total":79,"rows":[{"id":"2101.03961","title":"Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity","created":"2021-01-11","cited_by_count":3073,"area":"moe"},{"id":"2401.04088","title":"Mixtral of Experts","created":"2024-01-08","cited_by_count":1823,"area":"moe"},{"id":"2405.04434","title":"DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model","created":"2024-05-07","cited_by_count":1135,"area":"moe"},{"id":"2401.06066","title":"DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models","created":"2024-01-11","cited_by_count":867,"area":"moe"},{"id":"2502.16982","title":"Muon is Scalable for LLM Training","created":"2025-02-24","cited_by_count":428,"area":"moe"},{"id":"2201.05596","title":"DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale","created":"2022-01-14","cited_by_count":404,"area":"moe"},{"id":"2202.08906","title":"ST-MoE: Designing Stable and Transferable Sparse Expert Models","created":"2022-02-17","cited_by_count":399,"area":"moe"},{"id":"2406.11931","title":"DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence","created":"2024-06-17","cited_by_count":331,"area":"moe"},{"id":"2601.02780","title":"MiMo-V2-Flash Technical Report","created":"2026-01-06","cited_by_count":191,"area":"moe"},{"id":"2402.01739","title":"OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models","created":"2024-01-29","cited_by_count":187,"area":"moe"},{"id":"2411.02265","title":"Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencent","created":"2024-11-04","cited_by_count":90,"area":"moe"},{"id":"2601.07372","title":"Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models","created":"2026-01-12","cited_by_count":86,"area":"moe"},{"id":"2305.12281","title":"Lifelong Language Pretraining with Distribution-Specialized Experts","created":"2023-05-20","cited_by_count":66,"area":"moe"},{"id":"2404.07413","title":"JetMoE: Reaching Llama2 Performance with 0.1M Dollars","created":"2024-04-11","cited_by_count":57,"area":"moe"},{"id":"2405.03133","title":"Lory: Fully Differentiable Mixture-of-Experts for Autoregressive Language Model Pre-training","created":"2024-05-06","cited_by_count":37,"area":"moe"},{"id":"2505.12082","title":"Model Merging in Pre-training of Large Language Models","created":"2025-05-17","cited_by_count":37,"area":"moe"},{"id":"2509.09660","title":"Steering MoE LLMs via Expert (De)Activation","created":"2025-09-11","cited_by_count":26,"area":"moe"},{"id":"2502.19261","title":"Drop-Upcycling: Training Sparse Mixture of Experts with Partial Re-initialization","created":"2025-02-26","cited_by_count":24,"area":"moe"},{"id":"2406.12034","title":"Self-MoE: Towards Compositional Large Language Models with Self-Specialized Experts","created":"2024-06-17","cited_by_count":23,"area":"moe"},{"id":"2505.23854","title":"Revisiting Uncertainty Estimation and Calibration of Large Language Models","created":"2025-05-29","cited_by_count":23,"area":"moe"},{"id":"2401.02731","title":"Parameter-Efficient Sparsity Crafting from Dense to Mixture-of-Experts for Instruction Tuning on General Tasks","created":"2024-01-05","cited_by_count":22,"area":"moe"},{"id":"2407.01906","title":"Let the Expert Stick to His Last: Expert-Specialized Fine-Tuning for Sparse Architectural Large Language Models","created":"2024-07-02","cited_by_count":21,"area":"moe"},{"id":"2410.19034","title":"Mixture of Parrots: Experts improve memorization more than reasoning","created":"2024-10-24","cited_by_count":21,"area":"moe"},{"id":"2411.12364","title":"Ultra-Sparse Memory Network","created":"2024-11-19","cited_by_count":19,"area":"moe"},{"id":"2408.08274","title":"BAM! Just Like That: Simple and Efficient Parameter Upcycling for Mixture of Experts","created":"2024-08-15","cited_by_count":16,"area":"moe"},{"id":"2411.02433","title":"SLED: Self Logits Evolution Decoding for Improving Factuality in Large Language Models","created":"2024-11-01","cited_by_count":16,"area":"moe"},{"id":"2507.07024","title":"FlexOlmo: Open Language Models for Flexible Data Use","created":"2025-07-09","cited_by_count":15,"area":"moe"},{"id":"2501.10375","title":"DAOP: Data-Aware Offloading and Predictive Pre-Calculation for Efficient MoE Inference","created":"2024-12-16","cited_by_count":14,"area":"moe"},{"id":"2509.21892","title":"Elastic MoE: Unlocking the Inference-Time Scalability of Mixture-of-Experts","created":"2025-09-26","cited_by_count":11,"area":"moe"},{"id":"2512.23447","title":"Coupling Experts and Routers in Mixture-of-Experts via an Auxiliary Loss","created":"2025-12-29","cited_by_count":11,"area":"moe"},{"id":"2602.17004","title":"Arcee Trinity Large Technical Report","created":"2026-02-19","cited_by_count":11,"area":"moe"},{"id":"2509.16882","title":"Dynamic Expert Specialization: Towards Catastrophic Forgetting-Free Multi-Domain MoE Adaptation","created":"2025-09-21","cited_by_count":10,"area":"moe"},{"id":"2602.03560","title":"HySparse: A Hybrid Sparse Attention Architecture with Oracle Token Selection and KV Cache Sharing","created":"2026-02-03","cited_by_count":10,"area":"moe"},{"id":"2606.04438","title":"LoopMoE: Unifying Iterative Computation with Mixture-of-Experts for Language Modeling","created":"2026-06-03","cited_by_count":10,"area":"moe"},{"id":"2311.10768","title":"Memory Augmented Language Models through Mixture of Word Experts","created":"2023-11-15","cited_by_count":9,"area":"moe"},{"id":"2407.09816","title":"MaskMoE: Boosting Token-Level Learning via Routing Mask in Mixture-of-Experts","created":"2024-07-13","cited_by_count":9,"area":"moe"},{"id":"2405.15052","title":"Revisiting MoE and Dense Speed-Accuracy Comparisons for LLM Training","created":"2024-05-23","cited_by_count":8,"area":"moe"},{"id":"2504.02658","title":"MiLo: Efficient Quantized MoE Inference with Mixture of Low-Rank Compensators","created":"2025-04-03","cited_by_count":8,"area":"moe"},{"id":"2408.15901","title":"Nexus: Specialization meets Adaptability for Efficiently Training Mixture of Experts","created":"2024-08-28","cited_by_count":7,"area":"moe"},{"id":"2410.01610","title":"Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging","created":"2024-10-02","cited_by_count":6,"area":"moe"},{"id":"2505.09519","title":"PT-MoE: An Efficient Finetuning Framework for Integrating Mixture-of-Experts into Prompt Tuning","created":"2025-05-14","cited_by_count":6,"area":"moe"},{"id":"2510.02324","title":"Hallucination reduction with CASAL: Contrastive Activation Steering For Amortized Learning","created":"2025-09-25","cited_by_count":6,"area":"moe"},{"id":"2409.12210","title":"Mixture of Diverse Size Experts","created":"2024-09-18","cited_by_count":5,"area":"moe"},{"id":"2410.16077","title":"CartesianMoE: Boosting Knowledge Sharing among Experts via Cartesian Product Routing in Mixture-of-Experts","created":"2024-10-21","cited_by_count":5,"area":"moe"},{"id":"2503.05029","title":"Continual Pre-training of MoEs: How robust is your router?","created":"2025-03-06","cited_by_count":5,"area":"moe"},{"id":"2506.05767","title":"dots.llm1 Technical Report","created":"2025-06-06","cited_by_count":5,"area":"moe"},{"id":"2508.18756","title":"UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning","created":"2025-08-26","cited_by_count":5,"area":"moe"},{"id":"2510.16756","title":"End-to-end Listen, Look, Speak and Act","created":"2025-10-19","cited_by_count":5,"area":"moe"},{"id":"2602.01203","title":"Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse","created":"2026-02-01","cited_by_count":5,"area":"moe"},{"id":"2507.18671","title":"Innovator: Scientific Continued Pretraining with Fine-grained MoE Upcycling","created":"2025-07-24","cited_by_count":4,"area":"moe"},{"id":"2603.18297","title":"Path-Constrained Mixture-of-Experts","created":"2026-03-18","cited_by_count":4,"area":"moe"},{"id":"2605.06663","title":"EMO: Pretraining Mixture of Experts for Emergent Modularity","created":"2026-05-07","cited_by_count":4,"area":"moe"},{"id":"2412.15411","title":"Sparse Checkpointing for Fast and Reliable MoE Training","created":"2024-12-19","cited_by_count":3,"area":"moe"},{"id":"2601.21420","title":"ConceptMoE: Adaptive Token-to-Concept Compression for Implicit Compute Allocation","created":"2026-01-29","cited_by_count":3,"area":"moe"},{"id":"2602.00800","title":"JTok: On Token Embedding as another Axis of Scaling Law via Joint Token Self-modulation","created":"2026-01-31","cited_by_count":3,"area":"moe"},{"id":"2605.27358","title":"MobileMoE: Scaling On-Device Mixture of Experts","created":"2026-05-26","cited_by_count":3,"area":"moe"},{"id":"2504.15527","title":"Compass-V2 Technical Report","created":"2025-04-22","cited_by_count":2,"area":"moe"},{"id":"2505.21191","title":"Unveiling Instruction-Specific Neurons & Experts: An Analytical Framework for LLM's Instruction-Following Capabilities","created":"2025-05-27","cited_by_count":2,"area":"moe"},{"id":"2604.23108","title":"Mixture of Heterogeneous Grouped Experts for Language Modeling","created":"2026-04-25","cited_by_count":2,"area":"moe"},{"id":"2412.11834","title":"Wonderful Matrices: Combining for a More Efficient and Effective Foundation Model Architecture","created":"2024-12-16","cited_by_count":1,"area":"moe"},{"id":"2502.12455","title":"DSMoE: Matrix-Partitioned Experts with Dynamic Routing for Computation-Efficient Dense LLMs","created":"2025-02-18","cited_by_count":1,"area":"moe"},{"id":"2509.00679","title":"Router Upcycling: Leveraging Mixture-of-Routers in Mixture-of-Experts Upcycling","created":"2025-08-31","cited_by_count":1,"area":"moe"},{"id":"2601.17063","title":"FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge Devices","created":"2026-01-22","cited_by_count":1,"area":"moe"},{"id":"2609.40316","title":"Scaling Laws for Looped Mixture of Experts","created":"2026-09-30","cited_by_count":1,"area":"moe"},{"id":"2412.17009","title":"Expert Routing with Synthetic Data for Continual Learning","created":"2024-12-22","cited_by_count":0,"area":"moe"},{"id":"2509.23012","title":"MoE-PHDS: One MoE checkpoint for flexible runtime sparsity","created":"2025-09-27","cited_by_count":0,"area":"moe"},{"id":"2510.26182","title":"MossNet: Mixture of State-Space Experts is a Multi-Head Attention","created":"2025-10-30","cited_by_count":0,"area":"moe"},{"id":"2602.08818","title":"FlexMoRE: A Flexible Mixture of Rank-heterogeneous Experts for Efficient Federatedly-trained Large Language Models","created":"2026-02-09","cited_by_count":0,"area":"moe"},{"id":"2604.14246","title":"Awakening Dormant Experts:Counterfactual Routing to Mitigate MoE Hallucinations","created":"2026-04-15","cited_by_count":0,"area":"moe"},{"id":"2605.13247","title":"EMO: Frustratingly Easy Progressive Training of Extendable MoE","created":"2026-05-13","cited_by_count":0,"area":"moe"},{"id":"2606.12397","title":"Redesign Mixture-of-Experts Routers with Manifold Power Iteration","created":"2026-06-10","cited_by_count":0,"area":"moe"},{"id":"2608.29790","title":"HiVe: Beyond Static Prompts for Multitask Learning via Hierarchy-based Vertical Mixture-of-Experts","created":"2026-08-30","cited_by_count":0,"area":"moe"},{"id":"2609.00791","title":"Instella-MoE Technical Report","created":"2026-09-01","cited_by_count":0,"area":"moe"},{"id":"2609.28053","title":"Exact Quantile Balancing and Load-Error Injection for Mixture-of-Experts","created":"2026-09-23","cited_by_count":0,"area":"moe"},{"id":"2609.32398","title":"Function Over Form: Distributional Orthogonalization in Mixture-of-Experts with Replica Expert Mechanism","created":"2026-09-26","cited_by_count":0,"area":"moe"},{"id":"2609.36301","title":"MoRE: Scaling mixture of experts with hardware-aware low-rank routing","created":"2026-09-28","cited_by_count":0,"area":"moe"},{"id":"2610.02123","title":"Harnessing Domain Specialists in Multimodal Mixture-of-Experts for Efficient Adaptation","created":"2026-10-01","cited_by_count":0,"area":"moe"},{"id":"2610.11317","title":"DivMoE: Fine-Grained MoE Upcycling via Cross-Domain Expert Composition","created":"2026-10-08","cited_by_count":0,"area":"moe"},{"id":"2610.14074","title":"One Model, Multiple Compute Budgets: Multi-Ratio ConceptMoE for Flexible Compute Across Turns","created":"2026-10-08","cited_by_count":0,"area":"moe"}],"next_cursor":null}