ALTA:Compiler-BasedAnalysisofTransformers
We propose a new programming language called ALTA and a compiler that can map ALTA programs to Trans
首页 > 研究论文
-
-
LeveragingSkillsfromUnlabeledPriorDataforEfficientOnlineExploration
Unsupervised pretraining has been transformative in many supervised domains. However, applying such
-
WorldSimBench:TowardsVideoGenerationModelsasWorldSimulators
Recent advancements in predictive models have demonstrated exceptional capabilities in predicting th
-
TP-Eval:TapMultimodalLLMs'PotentialinEvaluationbyCustomizingPrompts
Recently, multimodal large language models (MLLMs) have received much attention for their impressive
-
TrainingFreeGuidedFlowMatchingwithOptimalControl
Controlled generation with pre-trained Diffusion and Flow Matching models has vast applications. One
-
Beyondposition:howrotaryembeddingsshaperepresentationsandmemoryinautoregressivetransfomers
Rotary Positional Embeddings (RoPE) enhance positional encoding in Transformer models, yet their ful
-
SPIRE:SynergisticPlanning,Imitation,andReinforcementLearningforLong-HorizonManipulation
Robot learning has proven to be a general and effective technique for programming manipulators. Imit
-
ExplainingBayesianNetworksinNaturalLanguageusingFactorArguments.Evaluationinthemedicaldomain
In this paper, we propose a model for building natural language explanations for Bayesian Network Re
-
KeyAlgorithmsforKeyphraseGeneration:Instruction-BasedLLMsforRussianScientificKeyphrases
Keyphrase selection is a challenging task in natural language processing that has a wide range of ap
-
GraphTeam:FacilitatingLargeLanguageModel-basedGraphAnalysisviaMulti-AgentCollaboration
Graphs are widely used for modeling relational data in real-world scenarios, such as social networks
-
Cross-lingualTransferofRewardModelsinMultilingualAlignment
Reinforcement learning with human feedback (RLHF) is shown to largely benefit from precise reward mo
-
BenchmarkingFoundationModelsonExceptionalCases:DatasetCreationandValidation
Foundation models (FMs) have achieved significant success across various tasks, leading to research
-
AIdrivenhealthrecommender
As AI emerged as highest valued technology, We used that to create a web application that makes a pa
-
FederatedTransformer:Multi-PartyVerticalFederatedLearningonPracticalFuzzilyLinkedData
Federated Learning (FL) is an evolving paradigm that enables multiple parties to collaboratively tra
-
DynamicSpectrumAccessforAmbientBackscatterCommunication-assistedD2DSystemswithQuantumReinforcementLearning
Spectrum access is an essential problem in device-to-device (D2D) communications. However, with the
-
OpticalGenerativeModels
Generative models cover various application areas, including image, video and music synthesis, natur
-
Closed-formmergingofparameter-efficientmodulesforFederatedContinualLearning
Model merging has emerged as a crucial technique in Deep Learning, enabling the integration of multi
-
MCUBERT:Memory-EfficientBERTInferenceonCommodityMicrocontrollers
In this paper, we propose MCUBERT to enable language models like BERT on tiny microcontroller units
-
ExpertFlow:OptimizedExpertActivationandTokenAllocationforEfficientMixture-of-ExpertsInference
Sparse Mixture of Experts (MoE) models, while outperforming dense Large Language Models (LLMs) in te
-
SimRAG:Self-ImprovingRetrieval-AugmentedGenerationforAdaptingLargeLanguageModelstoSpecializedDomains
Retrieval-augmented generation (RAG) enhances the question-answering (QA) abilities of large languag
