© 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. 35749
    Prev
    Next
    INDEX
    Explanations

    told you

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

    The neuron flags tokens that look like dialogue attributions or speaker labels (e.g. “B:”, “Right.”, “David says”), i.e. markers of a change in speaker in reported speech.

    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
    getTest
    -1.02
    hdmi
    -0.90
    мий
    -0.89
    Preface
    -0.88
     Zerst
    -0.86
     Prü
    -0.85
    нець
    -0.84
    ческие
    -0.84
    Sorry
    -0.84
    cartItem
    -0.83
    POSITIVE LOGITS
     preparar
    0.93
    monės
    0.88
    可能性
    0.87
    AutoScaleMode
    0.85
     gesprek
    0.81
    ścią
    0.81
    couverte
    0.81
     mentioning
    0.81
    enosti
    0.81
     speculating
    0.81
    Activations Density 0.028%

    No Known Activations