© 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. 22-GEMMASCOPE-RES-131K
    7. 66750
    Prev
    Next
    INDEX
    Explanations

    by or after

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

    This neuron specializes in detecting technical procedural‐step verbs that describe methods or instructions (e.g. “reached,” “sourced,” “opens,” “generated,” “back,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    sequently
    -0.85
    moved
    -0.78
    AGON
    -0.78
    weeney
    -0.76
     painel
    -0.76
     Subsequently
    -0.75
    Lastly
    -0.74
    chtete
    -0.74
    ierzch
    -0.73
    ∗
    -0.73
    POSITIVE LOGITS
     after
    1.80
     via
    1.68
     through
    1.56
    可以通过
    1.36
     thanks
    1.27
     благодаря
    1.27
     obtained
    1.22
    通过
    1.20
     from
    1.18
     by
    1.17
    Activations Density 0.271%

    No Known Activations