© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-RES-262K
    4. 5437
    Prev
    Next
    INDEX
    Explanations

    roles followed by descriptions

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

    The neuron detects words or phrases that name people by their role, occupation, title, or character label (e.g., professions and character descriptors).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    to
    0.32
    C
    0.31
    R
    0.31
    as
    0.31
    S
    0.31
    L
    0.30
    ↵↵
    0.29
    is
    0.29
    t
    0.28
    a
    0.28
    POSITIVE LOGITS
     who
    0.56
     الذين
    0.54
     ktorí
    0.52
     quien
    0.45
     kteří
    0.45
     quienes
    0.44
     जिन्होंने
    0.42
     którzy
    0.41
    who
    0.39
     extraordinaire
    0.39
    Activations Density 0.307%

    No Known Activations