© 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-3-27B-IT
    5. 5-GEMMASCOPE-2-TRANSCODER-262K
    7. 210372
    Prev
    Next
    INDEX
    Explanations

    This list doesn't immediately reveal a clear pattern related to digit '3' and its subsequent tokens. The logits are very diverse.Let's re-examine the MAX_ACTIVATING_TOKENS and TOKENS_AFTER_MAX_ACTIVATING_TOKEN.MAX_ACTIVATING_TOKENS: `3` (repeatedly)TOKENS_AFTER_MAX_ACTIVATING_TOKEN: `7`, `7`, `in`, `7`, `chip`, `,`, `7`, `A`The consistent '3' token in `MAX_ACTIVATING_TOKENS` is the strongest signal.the number 3

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_5_width_262k_l0_small_affine
    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
    나
    1.09
    0
    1.03
    به
    0.92
    스
    0.92
    보
    0.92
    Да
    0.91
    两国
    0.91
    কে
    0.90
    па
    0.88
    কিছু
    0.88
    POSITIVE LOGITS
    x
    1.47
    c
    1.46
    l
    1.36
    w
    1.35
    v
    1.32
    f
    1.30
    z
    1.16
    ל
    1.10
    ro
    1.08
    d
    1.06
    Activations Density 0.000%

    No Known Activations