© 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. 93510
    Prev
    Next
    INDEX
    Explanations

    question or list introducers

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

    The neuron fires on text that looks like a news‐style headline—especially proper names and the “X for Y over ‘…’” defamation-style constructions.

    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
     impressively
    -0.80
    崑
    -0.77
    あい
    -0.77
     EVERY
    -0.77
     Topography
    -0.76
     perpetually
    -0.75
     HUGE
    -0.75
    gania
    -0.75
     marvellous
    -0.75
     elevations
    -0.75
    POSITIVE LOGITS
     isKindOfClass
    0.98
    äldrar
    0.96
     valueForKey
    0.95
    wrights
    0.94
    roups
    0.93
    ressed
    0.93
    hess
    0.93
    ects
    0.93
     imprimé
    0.91
    lcccccc
    0.91
    Activations Density 0.008%

    No Known Activations