© 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. 71482
    Prev
    Next
    INDEX
    Explanations

    stochastic separation

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

    This neuron detects the numeric time‐code entries (the seconds and millisecond fields) used in subtitle cues (e.g. “00:00:32,332 --> 00:00:37,337”).

    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
     will
    -1.20
    colorado
    -1.17
     on
    -1.07
     if
    -1.05
     auf
    -1.04
     at
    -1.01
     has
    -0.98
    「
    -0.96
     as
    -0.96
     does
    -0.94
    POSITIVE LOGITS
    isman
    1.12
     Gehalt
    1.11
     девочки
    1.10
    ヴァン
    1.08
    glicher
    1.06
    无聊
    1.06
     продолжает
    1.06
     STARTED
    1.05
     courte
    1.04
    continues
    1.04
    Activations Density 0.000%

    No Known Activations