© 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-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 2893
    Prev
    Next
    INDEX
    Explanations

    words related to plans and actions

    oai_token-act-pair · gemini-2.0-flash

    terms related to catching

    np_token-act-pair-logits · gpt-4o-mini

    "catch"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    headless
    -0.52
    MetaObject
    -0.51
     StatelessWidget
    -0.51
    पया
    -0.51
     насељу
    -0.50
    GrantedAuthority
    -0.50
     straips
    -0.49
    MigrationBuilder
    -0.47
    ppins
    -0.47
     עלי
    -0.46
    POSITIVE LOGITS
     catch
    1.75
     catching
    1.66
     capture
    1.66
     caught
    1.54
     catches
    1.51
     Catch
    1.50
     capturing
    1.47
    catch
    1.45
     captured
    1.45
     Capture
    1.43
    Activations Density 2.406%

    No Known Activations