© 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. 8885
    Prev
    Next
    INDEX
    Explanations

    words referring to spatial orientation and component parts, especially verticality and horizontality

    oai_token-act-pair · gemini-2.0-flash

    terms related to vertical structure and orientation

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

    vertical and horizontal

    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
    SBATCH
    -0.82
     itſelf
    -0.81
     uſed
    -0.73
    ſelves
    -0.72
     Efq
    -0.70
     rapides
    -0.70
     createState
    -0.70
     morire
    -0.69
     énergé
    -0.69
     becauſe
    -0.69
    POSITIVE LOGITS
    EndGlobalSection
    0.67
     bin
    0.57
     directors
    0.52
     del
    0.52
     pro
    0.51
     por
    0.51
     ed
    0.50
     zu
    0.50
     Directors
    0.50
     da
    0.50
    Activations Density 4.834%

    No Known Activations