© 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. 7606
    Prev
    Next
    INDEX
    Explanations

    qualifying words like "highly" or "well" and sentences that include "if".

    oai_token-act-pair · gemini-2.0-flash

    conditional phrases indicating necessity

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

    if

    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
     JsonResult
    -0.47
    iją
    -0.44
     denen
    -0.44
     ceux
    -0.42
    <?
    -0.42
    기도
    -0.41
    mentaux
    -0.41
    Hauptartikel
    -0.39
    LETED
    -0.38
    navigationItem
    -0.38
    POSITIVE LOGITS
     necessary
    0.87
    Hentet
    0.79
     needed
    0.78
    necessary
    0.74
    tvguidetime
    0.74
    ThroughAttribute
    0.74
    TintMode
    0.71
    RegistryLite
    0.71
    tagext
    0.70
     Needed
    0.70
    Activations Density 2.211%

    No Known Activations