© 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. 85607
    Prev
    Next
    INDEX
    Explanations

    decisions, steps, and direction

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

    The neuron activates on abstract, thematic nouns (and related adjectives) that denote roles, relationships, directions, or conceptual “things” (e.g. role, direction, authority, steps, issues, relationship).

    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
    They
    -1.88
     for
    -1.69
     短发
    -1.55
    煽
    -1.51
    原材料
    -1.45
     Dieser
    -1.43
     início
    -1.42
    <td>
    -1.39
     just
    -1.37
    ７
    -1.37
    POSITIVE LOGITS
     you
    5.25
     we
    3.52
     it
    3.25
     they
    3.14
     that
    3.09
     your
    2.58
     personalizadas
    1.86
     podemos
    1.80
     our
    1.77
    袿
    1.76
    Activations Density 0.435%

    No Known Activations