© 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. 79034
    Prev
    Next
    INDEX
    Explanations

    spec followed by prefixes or suffixes

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

    The neuron responds to occurrences of the letter sequence “spec” inside words (e.g. in “specimen,” “speculative,” “Spectacular,” 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
     schild
    -0.78
    慣
    -0.75
     Ackerman
    -0.74
     Schloss
    -0.73
     hä
    -0.72
    עו
    -0.71
    ESTO
    -0.70
    rä
    -0.69
     melden
    -0.69
     depri
    -0.69
    POSITIVE LOGITS
     Spec
    1.58
    spec
    1.56
     spec
    1.56
    Spec
    1.45
     SPEC
    1.43
    SPEC
    1.31
     Speck
    1.27
    spectral
    1.20
     Spe
    1.08
     spé
    1.06
    Activations Density 0.018%

    No Known Activations