© 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. 6253
    Prev
    Next
    INDEX
    Explanations

    words that indicate temporal relationships or a sequence of events

    oai_token-act-pair · gemini-2.0-flash

    present participles and gerunds

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

    after

    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
     surla
    -0.63
    RenderAtEndOf
    -0.63
     wasn
    -0.57
     wasnt
    -0.55
    matchCondition
    -0.52
     tuvimos
    -0.52
     apparaat
    -0.50
    abetes
    -0.50
     tuve
    -0.49
     didnt
    -0.49
    POSITIVE LOGITS
     being
    0.66
    دانشنامهٔ
    0.65
     serving
    0.59
    serve
    0.59
     announcing
    0.58
     serve
    0.57
    ComponentName
    0.54
     regaining
    0.54
     attending
    0.53
     dropping
    0.53
    Activations Density 1.943%

    No Known Activations