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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 11473
    Prev
    Next
    INDEX
    Explanations

    a wide variety of context-specific terms in technical and scientific documents, including programming code, scientific papers, and configuration files.

    oai_token-act-pair · gemini-2.0-flash

    unrecognized or unusual characters

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

    long s spellings

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    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
    <bos>
    -0.82
    нен
    -0.28
    fa
    -0.28
    gel
    -0.28
    <eos>
    -0.28
    ↵↵
    -0.27
    mo
    -0.27
    ρυ
    -0.27
     -
    -0.26
    мо
    -0.26
    POSITIVE LOGITS
     Efq
    0.47
     ſeveral
    0.45
     Majefty
    0.44
    ^(@)
    0.42
     myſelf
    0.41
     Monfieur
    0.38
     becauſe
    0.38
     themſelves
    0.37
     pleaſure
    0.37
     leaſt
    0.37
    Activations Density 52.910%

    No Known Activations