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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32965
    Prev
    Next
    INDEX
    Explanations

    describing tone or quality

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

    The neuron is most responsive to unusually long tokens (i.e. long, multi‐character words).

    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
     ficción
    -1.08
     fictional
    -1.06
    tissa
    -1.05
    fiction
    -0.95
     Fiction
    -0.95
     FICTION
    -0.93
     Azevedo
    -0.90
    Fiction
    -0.87
     fic
    -0.87
    śl
    -0.85
    POSITIVE LOGITS
     grounded
    1.23
    grounded
    1.06
     episodic
    0.96
    mido
    0.94
    pantal
    0.94
     ști
    0.92
     Earn
    0.89
     Dn
    0.88
    OGN
    0.87
    noir
    0.86
    Activations Density 0.039%

    No Known Activations