© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 12547
    Prev
    Next
    INDEX
    Explanations

    days of the week and current events

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

    Neuron 4 activates on conversational interview language—especially second-person pronouns and question prompts directed at a guest (e.g. “I know you’ve been … what are you hearing so far?”).

    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
     précieux
    -1.64
     siempre
    -1.52
     zawsze
    -1.45
     всегда
    -1.45
    鹇
    -1.41
     可愛い
    -1.38
     péro
    -1.38
    kleber
    -1.34
     Gründe
    -1.34
     reele
    -1.34
    POSITIVE LOGITS
     Thursday
    1.63
     tonight
    1.59
     —
    1.53
     last
    1.52
     four
    1.52
     Tuesday
    1.51
    ٦
    1.50
     trying
    1.49
     Friday
    1.48
     Wednesday
    1.48
    Activations Density 0.026%

    No Known Activations