Unverified60% confidenceFactExact time
Anthropic published a paper on superposition in 2022, revealing the phenomenon of individual neurons simultaneously encoding multiple concepts.
2
Sources
60%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
Jacobian-Lens in Practice: Manually Reshaping LLM Behavior with J-Space
redditr/LocalLLaMA7/11/2026
Related Claims
UnverifiedAnthropic published the dictionary learning research Towards Monosemanticity in 2023, attempting to decompose polysemantic neurons into monosemantic features via sparse autoencoders.75% similarUnverifiedIn 2023, an Anthropic research team published a paper using sparse autoencoder technology to decompose the internal neuron activations of the Claude model into millions of interpretable conceptual features.73% similarUnverifiedThe Neural ODE paper published by Chen et al. at NeurIPS in 2018 formally established the connection between backpropagation and the adjoint method63% similarUnverifiedDai等人在2022年发表了论文《Knowledge Neurons in Pretrained Transformers》，通过定位和编辑特定神经元可以修改模型对特定事实的记忆62% similarUnverifiedThe Download, the daily newsletter under MIT Technology Review, disclosed this neuroscience experiment.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/493355API
curl https://kongchang.com/api/v1/knowledge/claims/493355MCP
get_claim(id=493355)