© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 4580
    Prev
    Next
    INDEX
    Explanations

    various words from the Russian and German languages alongside some English

    oai_token-act-pair · gemini-2.0-flash

    prefixes or fragments of words

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

    multiple languages

    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
    ſelves
    -1.13
     Efq
    -1.10
     houſe
    -1.09
     myſelf
    -1.08
    ſelf
    -1.06
     itſelf
    -1.04
     Jefus
    -1.04
     Anſ
    -1.03
     Theſe
    -1.03
     Houſe
    -1.02
    POSITIVE LOGITS
     para
    0.53
    rad
    0.50
     é
    0.49
     
    0.49
     s
    0.46
     pr
    0.45
     الج
    0.44
    સ
    0.44
     ra
    0.44
     pri
    0.43
    Activations Density 0.219%

    No Known Activations