What does this research mean for the field?

DeepLIFT, a method that decomposes neural network predictions by backpropagating activation differences relative to a reference state, reveals feature dependencies missed by other approaches and provides significant advantages over gradient-based attribution methods. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

April 10, 2017Open Access

Learning Important Features Through Propagating Activation Differences

View Full Paper

Ask AI

Bookmark

Authors

ASAvanti ShrikumarCharles Sturt University PGPeyton GreensideCollege of San Mateo AKAnshul KundajeStanford Medicine

Discussion

Member takes

Implication

Key Points

Key points are not available for this paper at this time.

Cite This Study

Shrikumar et al. (2017) studied this question.

synapsesocial.com/papers/6a08eb62817c69ba7be4b5b5 https://doi.org/10.48550/arxiv.1704.02685

View Full Paper

Ask AI

Bookmark

April 10, 2017Open Access

Learning Important Features Through Propagating Activation Differences

View Full Paper

Ask AI

Bookmark

Authors

ASAvanti ShrikumarCharles Sturt University PGPeyton GreensideCollege of San Mateo AKAnshul KundajeStanford Medicine

Discussion

Member takes

Implication

Key Points

Key points are not available for this paper at this time.

Cite This Study

Shrikumar et al. (2017) studied this question.

synapsesocial.com/papers/6a08eb62817c69ba7be4b5b5 https://doi.org/10.48550/arxiv.1704.02685

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark