© 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. 31-GEMMASCOPE-2-TRANSCODER-262K
    7. 49830
    Prev
    Next
    INDEX
    Explanations

    overthinking and worry

    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_31_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
    ﺙ
    0.39
    雾
    0.38
    fahren
    0.37
    abbam
    0.37
    Amber
    0.37
    proposition
    0.36
    yta
    0.36
     scoff
    0.36
     spol
    0.35
     pressured
    0.35
    POSITIVE LOGITS
     worry
    0.59
     걱
    0.55
     Worry
    0.55
     worrying
    0.49
     worried
    0.49
     трево
    0.48
    担忧
    0.46
     चिंता
    0.44
     worries
    0.44
     worrisome
    0.43
    Activations Density 0.096%

    No Known Activations