© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 130203
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens that represent numeric values or literal numbers in code and text (e.g. digits, numeric constants, literal counts).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    programming types and variables

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     that
    -1.53
     this
    -1.12
     anklicken
    -1.09
     when
    -1.08
     one
    -1.06
     manchas
    -1.05
     chré
    -1.02
     σημαν
    -0.99
     slav
    -0.99
     at
    -0.99
    POSITIVE LOGITS
    </strong>
    1.59
    </b>
    1.20
    </i>
    1.15
    ឵
    1.15
    精心
    1.13
    again
    1.12
    </u>
    1.09
     alten
    1.06
    竟然是
    1.06
    intptr
    1.05
    Activations Density 0.001%

    No Known Activations