© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72707
    Prev
    Next
    INDEX
    Explanations

    calculate outcomes

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron responds to procedural definition phrases that explain something as a residual—e.g. uses of “subtract … and what is left is …” to define a value.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ändert
    -0.84
     باشد
    -0.81
    をつけて
    -0.81
     rappresenta
    -0.80
     decidir
    -0.78
     świąte
    -0.78
    لاک
    -0.76
    豊富
    -0.76
    savevideo
    -0.76
     criaturas
    -0.75
    POSITIVE LOGITS
     remaining
    0.79
    FileManager
    0.77
     slika
    0.76
     therapeutics
    0.76
    fontSize
    0.74
    diagnostics
    0.74
     другие
    0.72
    hloromethane
    0.72
    お
    0.71
     各
    0.71
    Activations Density 0.030%

    No Known Activations