© 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. 6290
    Prev
    Next
    INDEX
    Explanations

    seemingly random numbers, punctuation, and letters (likely referencing codes or IDs)

    oai_token-act-pair · gemini-2.0-flash

    non-English or non-standard token patterns

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

    say "3"

    np_max-act-logits · gemini-2.0-flash

    digits

    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
    HasAnnotation
    -0.52
     Thursday
    -0.52
    AxisAlignment
    -0.51
    Thursday
    -0.50
    esson
    -0.49
    amera
    -0.48
    rugated
    -0.48
    ted
    -0.48
    boni
    -0.47
    alin
    -0.47
    POSITIVE LOGITS
    KURZBESCHREIBUNG
    0.65
    
    0.63
    المكان
    0.63
    Šaltiniai
    0.60
    providedIn
    0.59
    :✨
    0.59
    الدراسه
    0.58
     ainfi
    0.57
     GetEnumerator
    0.55
     leaſt
    0.55
    Activations Density 1.861%

    No Known Activations