© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 44-GEMMASCOPE-2-TRANSCODER-262K
    7. 3091
    Prev
    Next
    INDEX
    Explanations

    Thought

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_44_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
    ᇁ
    0.42
    ⢿
    0.41
     canoeing
    0.41
     Fleetwood
    0.40
    സിന്റെ
    0.40
    .|__
    0.40
    فيذ
    0.38
    ınar
    0.38
     fart
    0.38
    粡
    0.38
    POSITIVE LOGITS
    vill
    0.40
     ardu
    0.39
    oring
    0.38
    kiy
    0.38
    ordano
    0.37
    ordu
    0.37
     लाव
    0.36
     xạ
    0.36
    fully
    0.35
     Callback
    0.34
    Activations Density 0.001%

    No Known Activations