© 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. 34-GEMMASCOPE-RES-131K
    7. 57790
    Prev
    Next
    INDEX
    Explanations

    HTML tags, specific bacteria, countries

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

    commas in lists of technical items or geographic locations.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    This neuron activates on individual list‐item labels or names—that is, single tokens appearing as entries in enumerated or bullet‐style lists (e.g. HTML tag names, header names, country names, constant identifiers, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     jesús
    -1.31
     bancaria
    -1.22
    mvh
    -1.21
     playera
    -1.21
    According
    -1.21
     moreno
    -1.20
     olx
    -1.20
    jaket
    -1.20
    ＝＝＝＝
    -1.19
    気がする
    -1.18
    POSITIVE LOGITS
    <bos>
    1.11
     '/':
    1.00
    [@"
    0.99
    asca
    0.98
    ”—
    0.97
    ConstraintMaker
    0.97
    toThrow
    0.97
    vikle
    0.93
     ала
    0.93
     —
    0.92
    Activations Density 0.034%

    No Known Activations