© 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. 71219
    Prev
    Next
    INDEX
    Explanations

    explaining specific aims

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

    The neuron activates on numeric list markers and section‐heading tokens (i.e. digits and their following punctuation like “1.” or “:” that denote numbered aims or outline items).

    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
     also
    -1.45
     these
    -1.32
     it
    -1.30
     también
    -1.23
     those
    -1.17
     any
    -1.09
     their
    -1.09
     соціаль
    -1.09
    felizmente
    -1.09
     your
    -1.08
    POSITIVE LOGITS
    FONTE
    1.01
    inige
    0.94
    唉
    0.92
    PARTE
    0.91
     detest
    0.90
     daß
    0.90
     HAPPENED
    0.89
    жений
    0.89
    enfer
    0.88
    estr
    0.88
    Activations Density 0.047%

    No Known Activations