Faculty

Institution

Unit

Venue

Sham Kakade

Gordon McKay Professor of Computer Science and Professor of Statistics

Harvard University

Affiliations

Harvard John A. Paulson School of Engineering and Applied Sciences

Applied Mathematics

Computer Science

Electrical Engineering

Materials Science & Mechanical Engineering

Sham Kakade is currently serving as an Applied Mathematics Gordon McKay Professor of Computer Science and Professor of Statistics at Harvard University.

Honors

Create free account to unlock more features

Experiences

Create free account to unlock more features

Education

Create free account to unlock more features

Sham Kakade

Harvard University

Gordon McKay Professor of Computer Science and Professor of Statistics

You can click a keyword to filter the displayed items

machine learning computer science applied mathematics mathematical optimization data science reinforcement learning machine learning research artificial intelligence neuroscience deep learning natural language processing learning theory supervised learning computational optimization pattern recognition computational intelligence computer vision computational learning theory convex analysis statistical learning theory stochastic gradient descent computational statistics markov decision process online learning high-dimensional statistics knowledge discovery engineering optimization matrix theory information theory exploration v exploitation canonical correlation analysis policy gradient methods music science music information retrieval random matrix sparse representations stochastic optimization gaussian process dopamine active learning bandit feedback saddle points tensor decompositions global convergence least squares numerical algorithms information technology statistical methodology spectral theory unsupervised machine learning intelligent information processing reward systems gaussian analysis parallel learning matrix analysis music classification audio retrieval regularization (mathematics) neuroinformatics geometric learning algorithmic learning sample complexity spectral algorithm spherical gaussians online optimization efficient learning least squares regression short note strong convexity data privacy communication security document clustering dimensionality reduction neural science hidden markov model statistical pattern recognition

Video Presentations

Notable Scholarly Works

Notable Federal Grants

Collaborators

A New Perspective on Shampoo's Preconditioner. arXiv (Cornell University) (2024)

Morwani, DepenShapira, ItaiVyas, NikhilMalach, EranKakade, ShamJanson, Lucas

A Study on the Calibration of In-Context Learning. (2024)

Hanlin ZhangYiFan ZhangYaodong YuDhruv MadekaDean FosterEric XingHimabindu LakkarajuSham Kakade

sensor calibrationubiquitous learningmachine learninglearning theorycalibrationnatural language processingautonomous learningcognitive sciencein-context learningcomputer science

Color-Filter: Conditional Loss Reduction Filtering for Targeted Language Model Pre-Training. arXiv (Cornell University) (2024)

Brandfonbrener, DavidZhang, HanlinKirsch, AndreasSchwarz, Jonathan RichardKakade, Sham

machine learningconditional loss reductionlanguage modelsnatural language processingcomputer science

Datacomp-Lm: In Search of the Next Generation of Training Sets for Language Models. arXiv (Cornell University) (2024)

Li, JeffreyFang, AlexSmyrnis, GeorgiosIvgi, MaorJordan, MattGadre, SamirBansal, HritikGuha, EtashKeh, SedrickArora, KushalGarg, SaurabhXin, RuiMuennighoff, NiklasHeckel, ReinhardMercat, JeanChen, MayeeGururangan, SuchinWortsman, MitchellAlbalak, AlonBitton, YonatanNezhurina, MariannaAbbas, AmroHsieh, Cheng-YuGhosh, DhrubaGardner, JoshKilian, MaciejZhang, HanlinShao, RulinPratt, SarahSanyal, SunnyIlharco, GabrielDaras, GiannisMarathe, KalyaniGokaslan, AaronZhang, JieyuChandu, KhyathiNguyen, ThaoVasiljevic, IgorKakade, ShamSong, ShuranSanghavi, SujayFaghri, FartashOh, SewoongZettlemoyer, LukeLo, KyleEl-Nouby, AlaaeldinPouransari, HadiToshev, AlexanderWang, StephanieGroeneveld, DirkSoldaini, LucaKoh, Pang WeiJitsev, JeniaKollar, ThomasDimakis, Alexandros G.Carmon, YairDave, AchalSchmidt, LudwigShankar, Vaishaal

natural language processingstatisticsmachine learning

Eliminating Position Bias of Language Models: A Mechanistic Approach. arXiv (Cornell University) (2024)

Wang, ZiqiZhang, HanlinLi, XinerHuang, Kuan-HaoHan, ChiJi, ShuiwangKakade, Sham M.Peng, HaoJi, Heng

attentionmultimedia retrievalcausal attentioncomputer visionsemantic evaluationlanguage modelsnatural language processingvision language modelvisual groundingvisual question answeringlanguagecognitive sciencemachine learningposition biasnatural language generation

Matching the Statistical Query Lower Bound for K-Sparse Parity Problems With Stochastic Gradient Descent. arXiv (Cornell University) (2024)

Kou, YiwenChen, ZixiangGu, QuanquanKakade, Sham M.

mathematical optimizationnatural language processingstochastic gradient descentapplied mathematicsk-sparse parity problems

Scaling Laws in Linear Regression: Compute, Parameters, and Data. arXiv (Cornell University) (2024)

Lin, LicongWu, JingfengKakade, Sham M.Bartlett, Peter L.Lee, Jason D.

linear regressioneconometricslawstatisticsscaling analysisapplied mathematicsregression analysis

Superposed Decoding: Multiple Generations From a Single Autoregressive Inference Pass. arXiv (Cornell University) (2024)

Shen, EthanFan, AlanPratt, Sarah MPark, Jae SungWallingford, MatthewKakade, Sham M.Holtzman, AriKrishna, RanjayFarhadi, AliKusupati, Aditya

multiple generationscomputer scienceinference

TRANSCENDENCE: Generative Models Can Outperform the Experts That Train Them. arXiv (Cornell University) (2024)

Zhang, EdwinZhu, VincentSaphra, NaomiKleiman, AnatEdelman, Benjamin L.Tambe, MilindKakade, Sham M.Malach, Eran

computer sciencemachine learning

Universal Length Generalization With Turing Programs. arXiv (Cornell University) (2024)

Hou, KaiyingBrandfonbrener, DavidKakade, ShamJelassi, SamyMalach, Eran

natural language processingcomputer science

1
2
3
4
5
22
10 / page

Sham Kakade

Gordon McKay Professor of Computer Science and Professor of Statistics

Harvard University

Affiliations

Harvard John A. Paulson School of Engineering and Applied Sciences

Applied Mathematics

Computer Science

Electrical Engineering

Materials Science & Mechanical Engineering

Sham Kakade is currently serving as an Applied Mathematics Gordon McKay Professor of Computer Science and Professor of Statistics at Harvard University.

Honors

Create free account to unlock more features

Experiences

Create free account to unlock more features

Education

Create free account to unlock more features

Sham Kakade

Harvard University

Gordon McKay Professor of Computer Science and Professor of Statistics

You can click a keyword to filter the displayed items

Video Presentations

Notable Scholarly Works

Notable Federal Grants

Collaborators

A New Perspective on Shampoo's Preconditioner. arXiv (Cornell University) (2024)

Morwani, DepenShapira, ItaiVyas, NikhilMalach, EranKakade, ShamJanson, Lucas

A Study on the Calibration of In-Context Learning. (2024)

Hanlin ZhangYiFan ZhangYaodong YuDhruv MadekaDean FosterEric XingHimabindu LakkarajuSham Kakade

sensor calibrationubiquitous learningmachine learninglearning theorycalibrationnatural language processingautonomous learningcognitive sciencein-context learningcomputer science

Color-Filter: Conditional Loss Reduction Filtering for Targeted Language Model Pre-Training. arXiv (Cornell University) (2024)

Brandfonbrener, DavidZhang, HanlinKirsch, AndreasSchwarz, Jonathan RichardKakade, Sham

machine learningconditional loss reductionlanguage modelsnatural language processingcomputer science

Datacomp-Lm: In Search of the Next Generation of Training Sets for Language Models. arXiv (Cornell University) (2024)

natural language processingstatisticsmachine learning

Eliminating Position Bias of Language Models: A Mechanistic Approach. arXiv (Cornell University) (2024)

Wang, ZiqiZhang, HanlinLi, XinerHuang, Kuan-HaoHan, ChiJi, ShuiwangKakade, Sham M.Peng, HaoJi, Heng

Matching the Statistical Query Lower Bound for K-Sparse Parity Problems With Stochastic Gradient Descent. arXiv (Cornell University) (2024)

Kou, YiwenChen, ZixiangGu, QuanquanKakade, Sham M.

mathematical optimizationnatural language processingstochastic gradient descentapplied mathematicsk-sparse parity problems

Scaling Laws in Linear Regression: Compute, Parameters, and Data. arXiv (Cornell University) (2024)

Lin, LicongWu, JingfengKakade, Sham M.Bartlett, Peter L.Lee, Jason D.

linear regressioneconometricslawstatisticsscaling analysisapplied mathematicsregression analysis

Superposed Decoding: Multiple Generations From a Single Autoregressive Inference Pass. arXiv (Cornell University) (2024)

Shen, EthanFan, AlanPratt, Sarah MPark, Jae SungWallingford, MatthewKakade, Sham M.Holtzman, AriKrishna, RanjayFarhadi, AliKusupati, Aditya

multiple generationscomputer scienceinference

TRANSCENDENCE: Generative Models Can Outperform the Experts That Train Them. arXiv (Cornell University) (2024)

Zhang, EdwinZhu, VincentSaphra, NaomiKleiman, AnatEdelman, Benjamin L.Tambe, MilindKakade, Sham M.Malach, Eran

computer sciencemachine learning

Universal Length Generalization With Turing Programs. arXiv (Cornell University) (2024)

Hou, KaiyingBrandfonbrener, DavidKakade, ShamJelassi, SamyMalach, Eran

natural language processingcomputer science

1
2
3
4
5
22
10 / page

Similar Faculty Members

Stefano Ermon

Institution

Stanford University

machine learning

computer science

Alejandro Ribeiro

Institution

University of Pennsylvania

optimal design

swarm intelligence

Peter Bartlett

Institution

University of California, Berkeley

data science

learning classifier system

Inderjit Dhillon

Institution

University of Texas at Austin

big data

numerical linear algebra

Martin Wainwright

Institution

Massachusetts Institute of Technology

information theory

computational optimization

Yan Liu

Institution

University of Southern California

explainable machine learning

applied artificial intelligence

Nikolaos Sidiropoulos

Institution

University of Virginia

signal processing

computer science

Quanquan Gu

Institution

University of California, Los Angeles

ai architecture

intelligent computing

Han Liu

Institution

Northwestern University

machine learning

applied mathematics

Vahid Tarokh

Institution

Duke University

sequential analysis

applied artificial intelligence

Similar Faculty Members

Stefano Ermon

Institution

Stanford University

machine learning

computer science

Alejandro Ribeiro

Institution

University of Pennsylvania

optimal design

swarm intelligence

Peter Bartlett

Institution

University of California, Berkeley

data science

learning classifier system

Inderjit Dhillon

Institution

University of Texas at Austin

big data

numerical linear algebra

Martin Wainwright

Institution

Massachusetts Institute of Technology

information theory

computational optimization

Yan Liu

Institution

University of Southern California

explainable machine learning

applied artificial intelligence

Nikolaos Sidiropoulos

Institution

University of Virginia

signal processing

computer science

Quanquan Gu

Institution

University of California, Los Angeles

ai architecture

intelligent computing

Han Liu

Institution

Northwestern University

machine learning

applied mathematics

Vahid Tarokh

Institution

Duke University

sequential analysis

applied artificial intelligence