© 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. 22-GEMMASCOPE-RES-131K
    7. 128628
    Prev
    Next
    INDEX
    Explanations

    This neuron responds to moderately long, content-bearing words (e.g. uncommon nouns or adjectives) rather than short function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    comparing or adding more

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     seems
    -1.08
     seem
    -1.05
     seemingly
    -0.98
    ttino
    -0.97
    フォーム
    -0.95
     Seems
    -0.93
     parecen
    -0.92
     seemed
    -0.91
     начали
    -0.91
     or
    -0.91
    POSITIVE LOGITS
    もう少し
    1.27
    もっと
    1.25
     altrett
    1.00
     аналоги
    0.98
     гораздо
    0.98
     this
    0.96
    がこの
    0.94
    similar
    0.93
    はこの
    0.93
     elsewhere
    0.92
    Activations Density 0.076%

    No Known Activations