© 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. 14286
    Prev
    Next
    INDEX
    Explanations

    mentions of methods of analysis, quantification, experimentation, and results (seems to be geared towards something of a "scientific method").

    oai_token-act-pair · gemini-2.0-flash

    words related to analysis or evaluation

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

    analyzing, evaluating

    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
     виправивши
    -0.50
    momix
    -0.46
    foque
    -0.44
    TestClass
    -0.42
    eqn
    -0.40
    AccessException
    -0.40
     lassen
    -0.40
    zensionen
    -0.39
     voks
    -0.39
     Conferences
    -0.39
    POSITIVE LOGITS
     measuring
    0.92
    queryInterface
    0.85
     studying
    0.84
     assessing
    0.83
     characterizing
    0.82
     evaluating
    0.82
     мәкал
    0.82
     analyzing
    0.81
     investigating
    0.79
     analysing
    0.76
    Activations Density 9.238%

    No Known Activations