© 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. 26047
    Prev
    Next
    INDEX
    Explanations

    documentation and metadata

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

    The neuron fires on tokens that represent numeric identifiers or digit‐heavy labels in code (e.g., IDs, counts, sequences).

    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
    rootView
    -1.48
     during
    -1.44
     will
    -1.36
    大军
    -1.33
    ēm
    -1.32
    军团
    -1.28
     System
    -1.27
    作战
    -1.26
    īj
    -1.25
    alertDialog
    -1.24
    POSITIVE LOGITS
     voyant
    1.67
    ÈNE
    1.66
     búsquedas
    1.65
     عاشقانه
    1.59
    ଂ
    1.56
    頜
    1.43
     twar
    1.42
     Preferencias
    1.40
    ophanes
    1.40
     Bemerkungen
    1.38
    Activations Density 0.036%

    No Known Activations