© 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. 2505
    Prev
    Next
    INDEX
    Explanations

    political terms, especially names of parties, politicians, and election-related words, and it also seems to activate for the conjunction "and".

    oai_token-act-pair · gemini-2.0-flash

    terms related to political support

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

    backing

    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
     Efq
    -0.88
     Monfieur
    -0.82
     Houſe
    -0.78
     houſe
    -0.77
    AndEndTag
    -0.75
     sandero
    -0.74
     Sito
    -0.70
     pouvoit
    -0.68
     Jefus
    -0.68
     myſelf
    -0.67
    POSITIVE LOGITS
     backed
    1.15
     backing
    1.06
    backed
    0.96
    Backing
    0.91
    backing
    0.86
     tipped
    0.85
     Backing
    0.82
     tip
    0.82
     backs
    0.81
    back
    0.70
    Activations Density 1.124%

    No Known Activations