© 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. 76975
    Prev
    Next
    INDEX
    Explanations

    license or obtain

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

    The neuron fires most strongly on short promotional or item-heading phrases that tout a benefit or result (e.g. “让你躺着也赚钱,” “Finally we can,” “cheap prices to attract customers”), i.e. it flags key words in benefit-promise statements.

    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
     Paxton
    -0.91
     colheres
    -0.88
    ennial
    -0.86
    QtCore
    -0.83
    薙
    -0.81
    antaranya
    -0.80
    السكان
    -0.80
     Participate
    -0.79
    篥
    -0.78
    dicionado
    -0.77
    POSITIVE LOGITS
     successfully
    1.48
     success
    1.33
     succeed
    1.28
     successful
    1.11
     succeeded
    1.05
    ได้
    1.04
     obtain
    1.02
     achieve
    1.00
    成功
    0.97
     succeeds
    0.96
    Activations Density 0.006%

    No Known Activations