© 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. 118043
    Prev
    Next
    INDEX
    Explanations

    looking and admiring

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

    The neuron fires strongly on “content” words (nouns, verbs, adjectives)—especially longer or semantically rich tokens—while largely ignoring short function 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
     SUMMARY
    -0.92
    endence
    -0.91
     hears
    -0.84
     experimenta
    -0.84
     Rptr
    -0.83
     integrand
    -0.81
    ковые
    -0.81
    utivos
    -0.80
     bandera
    -0.80
     fordi
    -0.79
    POSITIVE LOGITS
     admiring
    2.95
     viewing
    2.77
     staring
    2.73
     admire
    2.42
     gazing
    2.33
     stare
    2.31
     gaze
    2.27
     admired
    2.16
     looking
    2.14
     stared
    2.03
    Activations Density 0.067%

    No Known Activations