google/neural-tangents

Issues

• Tag a release

  #202 opened 6 months ago by PhilipVinc
  3

• Why is the covariance a 2D matrix?

  #160 opened 2 years ago by LeonhardStorm
  3

• Error with batching

  #206 opened 4 months ago by HaozhenZhao
  0

• High Memory Usage on Infinite-Width NTK for GPU Only

  #204 opened 5 months ago by deoliveirajoshua
  1

• Transformer Block Implementation gives NotImplementedError

  #203 opened 6 months ago by esnvidia
  0

• How to use batch to gradient_descent_mse_ensemble ?

  #196 opened 8 months ago by yanglebupt
  1

• Calling the empirical kernel function with different parameters returns same result

  #201 opened 9 months ago by PhilippMisofCH
  0

• Question regarding the cookbook

  #200 opened a year ago by XinyiYS
  0

• stax.serial PSDness

  #195 opened a year ago by wang-fang-cv
  1

• Inefficient jacobian computation for embedding layers.

  #199 opened a year ago by mohamad-amin
  1

• NKT_mean output Nan, when the number of training sample is increased

  #198 opened a year ago by zhangbububu
  3

• NTK/NNGP behavior in the infinite regime when weights are drawn from Gaussians with high standard deviation

  #197 opened a year ago by tengandreaxu
  7

• NTK is not PD

  #193 opened a year ago by bayesfourrier
  0

• using stax.Cos(a=1.0, b=1.0, c=0.0) to get kernel from conv layer gives error

  #192 opened a year ago by bayesfourrier
  2

• Erf function goes beyond [-1,1]

  #191 opened a year ago by bangxiangyong
  2

• How to do Aggregate on a Graph whose nodes are all vectors

  #176 opened 2 years ago by SchenbergZY
  6

• How to analyze transductive node classification

  #151 opened 3 years ago by joneswong
  2

• pip install issues

  #190 opened a year ago by SamTov
  2

• How to compute the empirical after kernel?

  #189 opened a year ago by VMS-6511
  1

• How to obtain aleatoric uncertainty?

  #188 opened a year ago by bangxiangyong
  2

• predict.py raises an error

  #152 opened a year ago by zeinabsadeghian
  7

• Conflicting dependencies between Jax and neural-tangents

  #171 opened a year ago by snehjp2
  6

• eNTK implementation uses deprecated xla attribute

  #186 opened a year ago by shawn-im
  2

• Colab notebooks issue

  #187 opened a year ago by juan190199
  2

• Question regarding OOM issues

  #182 opened 2 years ago by MichaelMMeskhi
  3

• Question regarding lr in Neural Tangents Cookbook

  #184 opened 2 years ago by alicedimo
  0

• Report of Abuse of misinformation in scientific research

  #175 opened 2 years ago by yhyu13
  0

• Question: Relu Kernel Computation

  #180 opened 2 years ago by yCobanoglu
  3

• Question: Connection MLE "parametrized" GP in infinite Width Limit vs minimizing MSE "parametrized" Kernel in infinite Width

  #181 opened 2 years ago by yCobanoglu
  4

• The analytical output of GP can not fit the result of NNGP generated by the nt.predict.gp_inference

  #178 opened 2 years ago by Gengfu-He
  1

• Are the parameters of the elementwise nonlinear layers trainable?

  #174 opened 2 years ago by rishiedupalli
  4

• AttributeError: module 'neural_tangents' has no attribute 'utils'

  #173 opened 2 years ago by LeavesLei
  8

• Defining Own Kernel Function in Neural Tangents

  #172 opened 2 years ago by snehjp2
  1

• The different output dimensions in examples

  #170 opened 2 years ago by cycyyy
  2

• Cases where NTK and NNGP are very different

  #169 opened 2 years ago by YehonatanAvidan
  0

• Excessive memory consumption for deep networks

  #168 opened 2 years ago by jglaser
  2

• Feature masks do not get reduced in the kernel

  #159 opened 2 years ago by jglaser
  2

• Questions about memory consumption of infinitely wide NTK

  #166 opened 2 years ago by jasonli0707
  6

• Does neural-tangents work for custom layer?

  #165 opened 2 years ago by Shuhul24
  1

• Draw Phase Diagram for CNTK

  #164 opened 2 years ago by hhorace
  1

• Question: simple example poor performance, what am I doing wrong?

  #163 opened 2 years ago by mfouesneau
  5

• neural tangents vs flax

  #162 opened 2 years ago by po-oya
  3

• Neural tangents finite network gives different results compared with JAX

  #155 opened 3 years ago by NiuJ1ao
  1

• Question about training model

  #150 opened 3 years ago by Firegreat123
  2

• `nt.empirical_ntk_fn` and `nt.linearize` do not compute gradients past the linear head in a ViT

  #149 opened 3 years ago by refitwhisper
  2

• Memory profiling with JAX pprof

  #147 opened 3 years ago by celidos
  1

• Question: I can't release the memory in gpu during execute nt.empirical_ntk_fn

  #146 opened 3 years ago by kkeevin123456
  4

• NTK for Inputs on Hypersphere

  #143 opened 3 years ago by lkskstlr
  2

• nt.linearize induces a CUDA_ERROR_OUT_OF_MEMORY error

  #144 opened 3 years ago by RylanSchaeffer
  7

• Imports not working

  #139 opened 3 years ago by alicedimo
  2