Unverified50% confidenceFactExact time
Neural networks have highly nonlinear and distributed representations: the same concept may be spread across a large number of neurons, and the same neuron may also participate in multiple concepts.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe knowledge of a neural network is stored in a distributed manner across all weight parameters, and a single weight may participate in the extraction of multiple features simultaneously.79% similarVerifiedInterpretability research from MIT and Anthropic found that there are large numbers of polysemantic features inside models, with a single neuron potentially responding simultaneously to multiple seemingly unrelated semantic concepts.77% similarUnverifiedBackpropagation requires the network to have a clear layered structure and differentiable activation functions, which is difficult to map biologically onto the real learning mechanisms of neurons.76% similarVerifiedMechanistic interpretability attempts to understand what concepts each neuron and each layer's connections encode inside a neural network.76% similarUnverified神经网络层数越多、神经元越多，模型的表达能力越强，但也越容易出现过拟合75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1070948API
curl https://kongchang.com/api/v1/knowledge/claims/1070948MCP
get_claim(id=1070948)