© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    Side-by-side compare needs a wider screen.
    Open other featureClose compare
    PrevNext
    Close
    Circuit Sparsity
    No connected neurons data available
    Explanations

    initializes sum to zero (negative activations)

    python-code · gemini-2.5-flash-lite
    Auto-Interp Tools
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    16,384 prompts, 256 tokens each
    Dataset (Dashboard)
    neuronpedia/python-code-simplified
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     instruction
    -0.09
     instructions
    -0.08
    gram
    -0.08
    struction
    -0.07
    ies
    -0.07
    y
    -0.07
    structions
    -0.07
    Y
    -0.06
    mory
    -0.06
    yp
    -0.06
    Positive Logits
    ange
    0.06
    ust
    0.05
    ry
    0.05
     optimal
    0.05
     of
    0.05
     t
    0.04
    _of
    0.04
    urn
    0.04
    of
    0.04
    ove
    0.04
    Activations Density 87.055%

    No Known Activations

    PrevNext
    Close
    Circuit Sparsity
    No connected neurons data available
    Explanations

    dictionary literal with string keys (negative activations)

    python-code · gemini-2.5-flash-lite
    Auto-Interp Tools
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    16,384 prompts, 256 tokens each
    Dataset (Dashboard)
    neuronpedia/python-code-simplified
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     P
    -0.02
    P
    -0.02
    class
    -0.02
    p
    -0.01
    arg
    -0.01
     default
    -0.01
     raise
    -0.01
     index
    -0.01
    ast
    -0.01
     i
    -0.01
    Positive Logits
    alue
    0.06
    ariable
    0.06
    ariables
    0.05
    Ô
    0.05
    Æ
    0.05
    udok
    0.05
    mmon
    0.05
    isited
    0.05
    vai
    0.05
    Ú
    0.05
    Activations Density 99.988%

    No Known Activations