© 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-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 8121
    Prev
    Next
    INDEX
    Explanations

    let's break down or reach out

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

    phrases that invite collaborative action to begin or organize an explanation, especially conversational openings proposing to “break down” or “figure out” something together.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_40_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
    for
    0.41
    and
    0.39
    red
    0.38
    U
    0.38
    3
    0.37
    in
    0.36
    T
    0.36
    content
    0.36
    network
    0.35
    data
    0.34
    POSITIVE LOGITS
     partake
    0.35
    باً
    0.32
     postul
    0.30
     usurp
    0.30
    ིན་
    0.29
     ejercer
    0.29
    ྥ
    0.29
    ﱢ
    0.29
     disting
    0.29
     protagonista
    0.29
    Activations Density 0.340%

    No Known Activations