© 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. 26373
    Prev
    Next
    INDEX
    Explanations

    What is X about

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

    This neuron detects colloquial “aside” or transitional phrases built around “thing is” (e.g. “the funny thing is,” “the strange thing is,” “what really bakes your noodle is,” 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
    是否
    -0.96
     копия
    -0.94
    otheses
    -0.92
    籏
    -0.92
     идея
    -0.91
     описа
    -0.90
    뻤
    -0.89
     dras
    -0.89
     是否
    -0.89
     Hacer
    -0.88
    POSITIVE LOGITS
     this
    1.59
     even
    1.14
     is
    1.13
    Unlike
    1.08
     also
    0.99
    Ingrédients
    0.99
     Unlike
    0.96
     these
    0.94
    jacke
    0.92
     jedoch
    0.91
    Activations Density 0.042%

    No Known Activations