© 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. 10-GEMMASCOPE-RES-131K
    7. 128874
    Prev
    Next
    INDEX
    Explanations

    tunable compositions and properties

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

    The neuron fires on technical adjectives denoting adjustability or controllability—most strongly on “tunable” (and its variants) and related “controllable”-type 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_10/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
     Existen
    -1.70
     Actualmente
    -1.55
     stehenden
    -1.47
     Incluso
    -1.43
    </h2>
    -1.43
     Sólo
    -1.41
    cún
    -1.41
     dette
    -1.38
     Debido
    -1.37
     Años
    -1.37
    POSITIVE LOGITS
     and
    2.09
     how
    1.66
     people
    1.53
     a
    1.49
     an
    1.45
     никогда
    1.41
    ším
    1.40
    询
    1.30
    ﾚ
    1.28
     referrerpolicy
    1.27
    Activations Density 0.127%

    No Known Activations