© 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. 22-GEMMASCOPE-TRANSCODER-16K
    7. 6454
    Prev
    Next
    INDEX
    Explanations

    words that are generally considered to be negative

    oai_token-act-pair · gemini-2.0-flash

    scandals and gossip

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
    andır
    -0.52
     Rest
    -0.50
    assertThrows
    -0.49
    domen
    -0.48
    assertRaises
    -0.46
     Director
    -0.44
     النا
    -0.44
     Pin
    -0.44
    Rest
    -0.44
     Scoop
    -0.43
    POSITIVE LOGITS
    ThroughAttribute
    0.85
     CreateTagHelper
    0.79
    EndGlobalSection
    0.75
     resourceCulture
    0.69
    tvguidetime
    0.68
     oprot
    0.64
    parsedMessage
    0.63
     متعلقه
    0.62
    Comté
    0.62
    EDEFAULT
    0.62
    Activations Density 2.080%

    No Known Activations