© 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. 10-GEMMASCOPE-RES-131K
    7. 13669
    Prev
    Next
    INDEX
    Explanations

    let off, out, down, in

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

    The neuron responds to the particles in English phrasal verbs (e.g. “out,” “off,” “up,” “on”) when they occur as part of a verb phrase.

    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
    J
    -2.67
    You
    -2.64
    Y
    -2.63
    v
    -2.61
    H
    -2.56
    R
    -2.53
    W
    -2.50
    Just
    -2.42
    While
    -2.41
    Make
    -2.41
    POSITIVE LOGITS
    沵
    2.27
     nenhuma
    2.14
     complètement
    2.08
    ꪻ
    2.08
    ⬈
    2.05
    蝿
    1.97
     montrent
    1.94
    ského
    1.94
     estavam
    1.91
    
    1.91
    Activations Density 0.003%

    No Known Activations