pytorch/captum

Issues

• Why do captum's perturbation and IG treat input & target differently?

  #1456 opened 12 days ago by rbelew
  0

• BertForMultipleChoice works without captum, breaks with it?

  #1273 opened 12 days ago by rbelew
  1

• Target not provided when necessary, cannot take gradient with respect to multiple outputs?

  #1358 opened 14 days ago by rbelew
  1

• Where do you apply the chat template in the example?

  #1451 opened a month ago by wendlerc
  0

• "The attention mask and the pad token id were not set" during attribution

  #1449 opened a month ago by RiverGao
  0

• Failing tests on `aarch64-darwin`

  #1447 opened a month ago by drupol
  0

• [BC breaking change in torch] `weights_only` default flip for `torch.load`

  #1443 opened a month ago by mikaylagawarecki
  0

• Using IntegratedGradients to explain LSTM model

  #1306 opened 6 months ago by qishubo
  3

• 'EpsilonRule' object has no attribute 'relevance_output'

  #1363 opened 2 months ago by Evansika959
  2

• How much memory required for LLMGradientAttribution.attribute() on LLama2 tutorial?

  #1430 opened 2 months ago by rbelew
  0

• How to use it to explain the multi-task models?

  #1423 opened 2 months ago by Sarah-air
  0

• Convergence of LayerIntegratedGradients

  #1413 opened 2 months ago by pakaphholbig
  0

• DeepLIFT with Softmax - normalization error

  #1367 opened 2 months ago by AITheorem
  0

• visualizer.render not working in jupyter notebook (javascript error)

  #1313 opened 5 months ago by ghazizadehlab
  4

• Use meta-llama/Llama-3.2-3B-Instruct,get unexpected result.

  #1366 opened 2 months ago by Ningshiqi
  1

• Why is sigmoid activation for LRP not allowed?

  #1361 opened 3 months ago by CloseChoice
  0

• DeepLift on Transformers Based Text Classifiers

  #1357 opened 3 months ago by fismimosa
  0

• GradientShap needs internal_batch_size argument to avoid out-of-memory errors

  #1350 opened 3 months ago by princyok
  0

• Retain Gradients for Input Samples During Explainability Methods

  #1346 opened 3 months ago by PietroMc
  0

• Integrated Gradients - Higher Convergence Delta with more Steps?

  #1334 opened 4 months ago by Steven-Palayew
  0

• when there are many input features ,how to use the IntegratedGradients?

  #1312 opened 5 months ago by Sarah-air
  1

• Add support for Tuples of Modules in LayerLRP

  #1329 opened 4 months ago by Quoding
  0

• Can’t create tensors as shown in first Captum Titanic tutorial?!

  #1325 opened 4 months ago by rbelew
  0

• Negative Scores in Feature Ablation and Choosing Baselines for Non-Image Data

  #1304 opened 5 months ago by samiraat
  0

• LayerConductance not working with llama2

  #1311 opened 5 months ago by ThePuscher
  0

• tcav diabetic retinopathy

  #1310 opened 6 months ago by FilipUniBe
  0

• captum.concept.TCAV() custom classifier

  #1309 opened 6 months ago by FilipUniBe
  0

• Captum TCAV train.py sgd_train_linear_model() Where do the weights come from?

  #1308 opened 6 months ago by FilipUniBe
  0

• Integrated Gradients for Llama2 May Produce Unstable Explanations for Long Contexts

  #1297 opened 6 months ago by WYT8506
  0

• Adding IDGI to Captum

  #1284 opened 7 months ago by ShreeSinghi
  4

• Integrated Gradient without edge attribute

  #1296 opened 6 months ago by dongwei2001
  0

• TypeError: 'Concept' object is not iterable

  #1291 opened 7 months ago by pris0
  1

• Deconvoltion and KernelSHAP methods create uninformative attribution maps

  #1290 opened 7 months ago by ZetrextJG
  0

• Getting target dimension is not valid even though the output shape is shape in GRAD CAM .

  #1289 opened 7 months ago by chahatsagarmain
  0

• TCAV score computation doesn't preserve experimental set order

  #1285 opened 7 months ago by pris0
  0

• What is the proper shape of baselines for DeepLiftShap?

  #1283 opened 7 months ago by barteksad
  0

• Integrated Gradients for Frame-Level Classification Tasks

  #1282 opened 7 months ago by david-gimeno
  0

• For sentence classification using BERT, PAD token is used in IG/Deeplift?

  #1269 opened 9 months ago by lkqnaruto
  1

• NoiseTunnel on LayerIntegratedGradients Perturbs Input Instead of Attribution Target (RunetimeError)

  #1281 opened 7 months ago by EldadTalShir
  0

• Incorrect indexing for max pooling layers with DeepLIFT

  #1276 opened 8 months ago by jmschrei
  1

• Integrated gradient on a function of models

  #1279 opened 8 months ago by MarcPoulin1
  0

• Possible to modify initial LRP score to compute CLRP(contrastive lrp)?

  #1275 opened 8 months ago by 1058433796
  0

• Integrated Gradients for Llama2

  #1271 opened 8 months ago by oscarwzt
  0

• what is the meaning of contribution?

  #1266 opened 9 months ago by Dongximing
  1

• Uninitialized target_tokens Variable in LLMGradientAttribution.attribute Method

  #1261 opened 9 months ago by MSingh-CSE
  0

• Explainability for embeddings

  #1254 opened 9 months ago by sherwin97
  0

• Addition of Important Direction Gradient Integration (IDGI) framework

  #1246 opened 9 months ago by ShreeSinghi
  0

• ImportError for 'TextTemplateFeature' and 'TextTokenFeature' from 'captum.attr'

  #1242 opened 9 months ago by MSingh-CSE
  2

• Error using Integrated Gradients on MPS

  #1240 opened 10 months ago by achouliaras
  1

• AttributeError: 'NoneType' object has no attribute 'relevance_input'

  #1238 opened a year ago by Pei4579
  4