© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 613
    Prev
    Next
    INDEX
    Explanations

    elements related to structure and organization in written content

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Follows specific words or characters

    np_acts-logits-general · gemini-2.0-flash

    regularly. always. yearly. always

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ProtoMessage
    -0.47
    enderror
    -0.42
    ongono
    -0.35
    dafx
    -0.35
    uzzy
    -0.33
    Aos
    -0.33
    autoIncrement
    -0.33
    mær
    -0.32
    prio
    -0.32
    aarrggbb
    -0.32
    POSITIVE LOGITS
     regularly
    0.77
    毎回
    0.74
    ALWAYS
    0.69
     yearly
    0.67
     always
    0.67
     siempre
    0.64
     annually
    0.64
     ALWAYS
    0.63
     sempre
    0.63
    always
    0.63
    Activations Density 0.703%

    No Known Activations