Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 8660
    Prev
    Next
    INDEX
    Explanations

    references to "content" in various contexts

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

    "Content" or its variations

    np_acts-logits-general · gemini-2.0-flash

    content analysis and validity

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     IBOutlet
    -0.82
     виправивши
    -0.65
    Tikang
    -0.64
     @"";
    -0.61
    }\]
    -0.60
     capuche
    -0.59
    JspWriter
    -0.58
    ఔ
    -0.58
     Thayer
    -0.58
    ſelf
    -0.58
    POSITIVE LOGITS
     Content
    0.81
     content
    0.78
     contents
    0.77
     setContent
    0.77
     getContent
    0.75
    setContent
    0.73
     CONTENT
    0.68
     ujednoznacz
    0.68
     material
    0.66
    getContent
    0.66
    Activations Density 0.070%

    No Known Activations