© 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. 34-GEMMASCOPE-RES-131K
    7. 65079
    Prev
    Next
    INDEX
    Explanations

    behind the scenes

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

    The neuron fires on news‐style reportage markers—things like headlines, section titles, and reporting verbs or key story words (“said,” “tells,” “interview,” “informed,” 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
    preduce
    -0.86
     Professor
    -0.73
     Sole
    -0.70
    urbation
    -0.69
    rpc
    -0.68
     professor
    -0.68
     doktor
    -0.67
     peng
    -0.66
    ptor
    -0.66
     bunu
    -0.66
    POSITIVE LOGITS
     backstage
    6.91
    后台
    2.97
     Back
    2.39
    Back
    2.33
    back
    2.11
    BACK
    2.08
     BACK
    2.03
     behind
    2.02
     backend
    1.98
     onstage
    1.91
    Activations Density 0.064%

    No Known Activations