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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 15346
    Prev
    Next
    INDEX
    Explanations

    phrases focused on the nature of horror and its elements, emphasizing the importance of subtlety and depth over mere shock value

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

    "about"

    np_acts-logits-general · gemini-2.0-flash

    the concept of about

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    parsedMessage
    -0.71
     يتيمه
    -0.58
     betweenstory
    -0.51
    DoubleQuotes
    -0.49
     שוליים
    -0.49
    RenderAtEndOf
    -0.47
    stateMutability
    -0.47
    vább
    -0.47
    EDEFAULT
    -0.45
     useAppContext
    -0.45
    POSITIVE LOGITS
     about
    1.50
     tentang
    1.06
    about
    1.06
    About
    0.96
     ABOUT
    0.95
     About
    0.95
    关于
    0.89
     acerca
    0.89
     tungkol
    0.87
    關於
    0.85
    Activations Density 0.168%

    No Known Activations