© 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. 9815
    Prev
    Next
    INDEX
    Explanations

    words relating to negative emotions and states, combined with intensifiers.

    oai_token-act-pair · gemini-2.0-flash

    dark themes and negative attributes

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

    negative adjectives

    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
     AssemblyCulture
    -0.88
     étoient
    -0.73
     الحره
    -0.70
     виправивши
    -0.68
     étoit
    -0.66
     referenties
    -0.66
     rempliss
    -0.65
    yorsunuz
    -0.63
     avoient
    -0.63
    enterOuterAlt
    -0.63
    POSITIVE LOGITS
     time
    0.54
     BoxFit
    0.49
     decision
    0.49
     things
    0.47
     but
    0.47
     behavior
    0.47
     pred
    0.47
    ade
    0.46
     numerus
    0.46
    xbb
    0.46
    Activations Density 2.305%

    No Known Activations